The final event of the “Learning About Your Genome” series aims to increase awareness of genetics in the African-American community.

Engaging the Community to Talk About Genetics

Sequencing an individual's genome once and re-using it as needed for different clinical questions over time could reduce costs, but genomic data is incredibly sensitive and misuse of its contents could have far-reaching consequences for individuals and ...

Geneial Receives New NIH Grant For Private Pharmacogenomics Workflow

Kasa Living, an S.F.-based provider of flexible accommodations for travelers, raised $70 million in Series C funding, Axios reports. Citi Ventures and FirstMark Capital co-led, and were joined by New York Life Ventures, Fireside Investments and insiders RET Ventures, Zigg Capital and Ribbit Capital.

Daily Digest: Cal plans for new 'innovation zone'; Splunk cuts jobs

Many hidden genetic variations can be detected with Chameleolyser, a new method developed in Nijmegen. The information is already yielding new patient diagnoses and may also lead to the discovery of as yet unknown disease genes,

New method reveals hidden genetic variations

Investigators have published and presented exciting clinical results in dozens of treated patients with sickle cell disease.

CRISPR Prime Time? FDA Advisory Committee Evaluates Vertex Exa-Cel Therapy for Sickle Cell Disease

A team of paleontologists, evolutionists and geoscientists affiliated with several entities in Germany has extracted Ice Age woolly rhino DNA from fossilized excrement samples (coprolites) found in a cave in Germany.

Ancient hyena droppings reveal genome of Ice Age woolly rhino

Newly formed partnership enables clinical genomics laboratories to move from sample to evidence-backed analysis to actionable clinical reports using a push-button solutionWASHINGTON--(BUSINESS WIRE)--#AI--Fabric Genomics,

Fabric Genomics and DNAnexus Team Up to Improve Scale and Speed of Data Analysis for Genomic Medicine

Poultry scientists with the Arkansas Agricultural Experiment Station are working to decrease the time and cost to diagnose an emerging disease affecting the turkey industry.Streptococcus gallolyticus,

Turkey disease being researched in Arkansas

As far as any concern for what’s called “off-target effects,” meaning the potential unwanted or adverse alterations to the genome that could accidentally happen in this process and cause cancer or other problems down the road, Dr. Daniel Bauer ...

FDA considers first CRISPR gene editing treatment that may cure sickle cell

The investment follows an earlier $23 million Series B funding round and will support the company's bioinformatics diagnostic tool development goals.

Genome Insight Secures Another $23M in Second Series B Round

Genome

genome

Why does the mind stay sharper than muscles as age advances? Why do professional athletes in several sports retire after 30? What is the difference between genes expressed elsewhere in the body and those expressed in the brain, by neurons? Why does muscle mass decrease after 30? If the mind controls muscle movement through life, does the mind have a role in muscle growth? An offspring does not inherit the language of the parent, but an offspring inherits anatomic features of the parent. What is the difference between the mechanism for language and those for the features of the parent? Both are made possible by genes, but one can be said to be indirect and the other say direct. Genes are present in all cells. They act as engines--with instructions--to make determinations. They work in muscle cells. They work across several cells and are expressed across functions. However, what is the relationship between when genes are expressed in several functions and when the brain is said to regulate those functions? Why is it possible that nurture for an offspring can result in huge differences from the parent, even if physical features are similar? What is the relationship between genes and the mind? Genes contain instructions, but it is postulated that they are almost totally regimented. Many of their instructions, conceptually, are not that flexible, so what they have to express is almost the same way it has to be expressed, without much changes--so to speak. It is theorized that what makes several functions--or options--possible is the mind. This means that genes are near fixed, but the mind runs the flexibility. A second difference is that the ability for the mind to give and receive information is greater than genes, that mostly send out instructions, though they receive information, it is not often as much as the mind, conceptually. Also, for what genes control directly and what the mind does too, directly, how do genes and the mind work together. What is the human mind? The mind is not genes nor neurons. The candidate for the human mind, conceptually, are the electrical and chemical signals of neurons, with their interactions and features in sets, in clusters of neurons across the central and peripheral nervous systems. Sets of signals have configurations [or arrangements] for which the basis of functions is held. This means that a smell, a sound, and so forth are all configurations of signals. Configuration templates arrive by the instruction of genes, but the contents are made by signals. However, there are several cases, conceptually, where genes are responsible for situations of the mind by anomalies in templates. This means that some individuals that have certain susceptibilities may have a template-induced configuration problem, such that when the problem takes place, with signals, it is at the error of genes. Simply, genes make it possible to have templates for which signals can build configurations. These templates--for several functions are generally similar--but anomalies exist within some, which may also be inherited, responsible for some kinds of addictions or mind disorders. A question is that for some of these template problems, are they from scratch or are they activated at a stage during development? Since genes are involved across cells, for development through life, it is possible to theorize that they also play roles in some of the templates for the configurations of mind, through adulthood, in many ways. This means that there could be continuous instruction passages about what templates should be, for certain functions, so that signals can then build configurations into them. In summary, genes can be said to make carvings, while signals enable a fill for those, as configurations. How Do Genes Work With The Mind? The mind is the front face of interaction with the world after birth, though the mind has already formed, conceptually before birth, it has no direct interface with the world. The mind becomes the operational source for many functions. Though genes in cells send out growth instructions, shaping cells for tissues, organs, and systems with timelines and extents, it is signals that operate at availability, while genes build availability. For example, the hand at any growth stage is by the genes, but its control or operation is by the mind. Genes can be said to have caps, theoretically, while the mind does not have caps. Even though genes lay out the possibility for what the mind can fill or build configurations for, the mind seems limitless with what it can learn, what it can be trained to do, though within the capacity of genes, but far beyond the allowance that genes have. It is theorized that a reason why sports retirement is common after 30, for many athletes, is that the stretch of genes becomes capped for certain activities around that age, even though the mind is still as sharp. Genes are also theorized to be responsible for aging, since their lack of continuous instruction for growth, unlike previously, may result in a peak then decline, by cells, of the tissue or organ. However, the mind, in use, often make it possible to keep functions active, even when growth by genes have stalled, conceptually. Mind and genes work closely together, but there is a lot that the mind decides. For example, what is the role of the mind in diseases? How do they relate to genes? Even though there are genetic anomalies responsible for some conditions, how does the mind factor? It is possible to use a conceptual model for the mind towards mapping the reach of genes against diseases. There is a recent feature, Gene therapy protects against motor neuron disease in rats, stating that, "University of Wisconsin–Madison researchers targeting a group of hereditary neurodegenerative diseases have found success using a gene therapy treatment in an animal model. The approach, which uses CRISPR-Cas9 genome editing technology, offers a unique and promising strategy that could one day treat rare but debilitating motor neuron diseases in humans. Hereditary spastic paraplegia, or HSP, is a group of movement disorders that cause progressive weakness and stiffness in the legs of people with certain inherited genetic mutations. The rare disorders usually lead to physical limitations and use of a wheelchair. Studying disease processes in animal models is an indispensable part of developing and testing any new treatments before they’re offered to humans. But scientists had historically struggled to replicate HSP’s symptoms and disease progression in animal models." There is another recent story, Aging-related genomic culprit found in Alzheimer’s disease – WashU Medicine, stating that, " Researchers at Washington University School of Medicine in St. Louis have developed a way to capture the effects of aging in the development of Alzheimer’s disease. They have devised a method to study aged neurons in the lab without a brain biopsy, an advancement that could contribute to a better understanding of the disease and new treatment strategies. The scientists transformed skin cells taken from patients with late-onset Alzheimer’s disease into brain cells called neurons. Late-onset Alzheimer’s develops gradually over many decades and only starts to show symptoms at age 65 or older. For the first time, these lab-derived neurons accurately reproduced the hallmarks of this type of dementia, including the amyloid beta buildup, tau protein deposits and neuronal cell death. By studying these cells, the researchers identified aspects of cells’ genomes — called retrotransposable elements, which change their activity as we age — in the development of late-onset Alzheimer’s disease. The findings suggest new treatment strategies targeting these factors." Why does the mind stay sharper than muscles as age advances? Why do professional athletes in several sports retire after 30? What is the difference between genes expressed elsewhere in the body and those expressed in the brain, by neurons? Why does muscle mass decrease after 30? If the mind controls muscle movement through life, does the mind have a role in muscle growth? An offspring does not inherit the language of the parent, but an offspring inherits anatomic features of the parent. What is the difference between the mechanism for language and those for the features of the parent? Both are made possible by genes, but one can be said to be indirect and the other say direct. Genes are present in all cells. They act as engines--with instructions--to make determinations. They work in muscle cells. They work across several cells and are expressed across functions. However, what is the relationship between when genes are expressed in several functions and when the brain is said to regulate those functions? Why is it possible that nurture for an offspring can result in huge differences from the parent, even if physical features are similar? What is the relationship between genes and the mind ? What is the relationship between genes and the mind Genes contain instructions, but it is postulated that they are almost totally regimented. Many of their instructions, conceptually, are not that flexible, so what they have to express is almost the same way it has to be expressed, without much changes--so to speak. It is theorized that what makes several functions--or options--possible is the mind. This means that genes are near fixed, but the mind runs the flexibility. A second difference is that the ability for the mind to give and receive information is greater than genes, that mostly send out instructions, though they receive information, it is not often as much as the mind, conceptually. Also, for what genes control directly and what the mind does too, directly, how do genes and the mind work together. What is the human mind? The mind is not genes nor neurons. The candidate for the human mind, conceptually, are the electrical and chemical signals of neurons, with their interactions and features in sets, in clusters of neurons across the central and peripheral nervous systems. Sets of signals have configurations [or arrangements] for which the basis of functions is held. This means that a smell, a sound, and so forth are all configurations of signals. Configuration templates arrive by the instruction of genes, but the contents are made by signals. templates However, there are several cases, conceptually, where genes are responsible for situations of the mind by anomalies in templates. This means that some individuals that have certain susceptibilities may have a template-induced configuration problem, such that when the problem takes place, with signals, it is at the error of genes. templates. template-induced Simply, genes make it possible to have templates for which signals can build configurations. These templates--for several functions are generally similar--but anomalies exist within some, which may also be inherited, responsible for some kinds of addictions or mind disorders. A question is that for some of these template problems, are they from scratch or are they activated at a stage during development? Since genes are involved across cells, for development through life, it is possible to theorize that they also play roles in some of the templates for the configurations of mind, through adulthood, in many ways. templates This means that there could be continuous instruction passages about what templates should be, for certain functions, so that signals can then build configurations into them. templates In summary, genes can be said to make carvings, while signals enable a fill for those, as configurations . while signals enable a fill for those, as configurations How Do Genes Work With The Mind? How Do Genes Work With The Mind? The mind is the front face of interaction with the world after birth, though the mind has already formed, conceptually before birth, it has no direct interface with the world. The mind becomes the operational source for many functions. Though genes in cells send out growth instructions, shaping cells for tissues, organs, and systems with timelines and extents, it is signals that operate at availability, while genes build availability. For example, the hand at any growth stage is by the genes, but its control or operation is by the mind. Genes can be said to have caps, theoretically, while the mind does not have caps. Even though genes lay out the possibility for what the mind can fill or build configurations for, the mind seems limitless with what it can learn, what it can be trained to do, though within the capacity of genes, but far beyond the allowance that genes have. It is theorized that a reason why sports retirement is common after 30, for many athletes, is that the stretch of genes becomes capped for certain activities around that age, even though the mind is still as sharp. Genes are also theorized to be responsible for aging, since their lack of continuous instruction for growth, unlike previously, may result in a peak then decline, by cells, of the tissue or organ. However, the mind, in use, often make it possible to keep functions active, even when growth by genes have stalled, conceptually. Mind and genes work closely together, but there is a lot that the mind decides. For example, what is the role of the mind in diseases? How do they relate to genes? Even though there are genetic anomalies responsible for some conditions, how does the mind factor? It is possible to use a conceptual model for the mind towards mapping the reach of genes against diseases. There is a recent feature, Gene therapy protects against motor neuron disease in rats , stating that, "University of Wisconsin–Madison researchers targeting a group of hereditary neurodegenerative diseases have found success using a gene therapy treatment in an animal model. The approach, which uses CRISPR-Cas9 genome editing technology, offers a unique and promising strategy that could one day treat rare but debilitating motor neuron diseases in humans. Hereditary spastic paraplegia, or HSP, is a group of movement disorders that cause progressive weakness and stiffness in the legs of people with certain inherited genetic mutations. The rare disorders usually lead to physical limitations and use of a wheelchair. Studying disease processes in animal models is an indispensable part of developing and testing any new treatments before they’re offered to humans. But scientists had historically struggled to replicate HSP’s symptoms and disease progression in animal models." Gene therapy protects against motor neuron disease in rats There is another recent story, Aging-related genomic culprit found in Alzheimer’s disease – WashU Medicine , stating that, " Researchers at Washington University School of Medicine in St. Louis have developed a way to capture the effects of aging in the development of Alzheimer’s disease. They have devised a method to study aged neurons in the lab without a brain biopsy, an advancement that could contribute to a better understanding of the disease and new treatment strategies. The scientists transformed skin cells taken from patients with late-onset Alzheimer’s disease into brain cells called neurons. Late-onset Alzheimer’s develops gradually over many decades and only starts to show symptoms at age 65 or older. For the first time, these lab-derived neurons accurately reproduced the hallmarks of this type of dementia, including the amyloid beta buildup, tau protein deposits and neuronal cell death. By studying these cells, the researchers identified aspects of cells’ genomes — called retrotransposable elements, which change their activity as we age — in the development of late-onset Alzheimer’s disease. The findings suggest new treatment strategies targeting these factors." Aging-related genomic culprit found in Alzheimer’s disease – WashU Medicine

Genes or Mind? Which is More Flexible? Roles in Aging, Sports Retirement and Growth

As the world transforms rapidly, especially after COVID-19, traditional institutions change faster too. Let’s take a look at the recent stats: In 2020, the number of e-commerce transactions made with digital or mobile wallets reached and by 2024 this number is expected to be over 50 percent. What does it have to do with neobanks or EMIs? roughly 45 percent They are called “challengers” for a reason and they are to take a share from traditional banks indeed. The venture capital in this segment is proves this statement. What is the nature of neobanks, what’s their place in fintech, and how businesses may need them – let’s find out in a minute. counted by billions What is a neobank? Neobank is a term for banks that exist only in digital form, as an app, web app, or other software solution, and doesn’t have any physical branches. It can work in several jurisdictions and is usually not limited to a single country or region. Sometimes “neobank” is used as an umbrella term for all apps with financial functional, like money transfers, but that’s not correct. Since not all financial apps have bank licenses or partnerships with physical banks, technically they can’t be considered as neobanks. Some neobanks to consider for businesses: (has EMI and bank license), . , N26 Revolut Starling Bank What is EMI? Electronic money institutions (EMIs) are certified to open e-wallets to their customers and have the right to issue electronic money. Although it looks very much like mobile banking or neobanks, it is not. EMIs can also potentially engage in payment services, cash withdrawal and deposit services, payment executions services, remittance services, direct debit or credit transfers, account information services, and payment initiation services. This list of “complementary” services into which the EMI can potentially engage strongly depends on the type of authorization as certain activities may be restricted. For instance, their main difference from banks is that they can't lend or invest clients’ money, and the sums of money over 100 000 EUR stored with them are not insured. If neobanks and EMI’s don’t have a banking license, they partner with physical banks. This allows for offering full banking services to their clients: checking, creating savings accounts, loans, and issuing credit cards. But their requirements for opening an account are usually less strict. Their main source of revenue comes from interest rates and ATMs out-of-networks. Some EMI’s to consider for businesses are , , , . Genome Fire Bunq Monzo Let’s pretend you make nice clothes and want to open an online boutique to sell them from your workshop. You take photos of all your items, set prices, write descriptions, and buy a domain for your website. Yet, you need a payment processing provider and an account to collect the payments and pay for website maintenance, marketing, etc. And here's where neobank and EMI come: you can open a merchant account and manage all these procedures online, usually faster than with a regular bank. It doesn’t mean that you don’t need to provide any documents or go through a verification process, but in a neobank, this process will be significantly faster. All the documents can be provided online and verified in 48-72 hours, while at a regular bank this procedure takes longer. As a result, you launch your business much faster and pay more attention not to paperwork, but to actual work - providing your customers with your high-quality things. Now that you get the general idea of neobanks, let’s proceed to some details. Pros and cons of neobanks and EMIs Neobanks seem like a good alternative to traditional banks and electronic wallets. Via them, customers can easily perform most of their daily financial operations. But does it mean that all they are as good for one people as for the others? Let’s take a look at their upsides and downsides. Advantages Agility. Being entirely digital products, neobanks follow the recent trends and often adopt innovations. Also, in case a problem with the solution arises, this allows the developers for fixing them quickly. Fast registration. One of the most valued advantages of digital banks is that you can open your account in up to 3 days and submit all necessary documents online. Multicurrency accounts. Traditional banks usually provide separate accounts for different currencies, while neobanks allow for creating multi-currency accounts. Fees and rates. As neobank owners spend less money on their maintenance, this results in lower fees for their clients and higher interest rates. Also, they do their best to make the terms of usage as transparent as possible. Finance education. Besides finance operations, these apps frequently offer their users finance management features like saving programs, budgeting, and finance management. Disadvantages Limited free services. The majority of neobanks offer free services, but this is a basic list and you might need more for your business. In this case, you can have them in a paid plan. However, even if you need just one service from their extra plan, you still pay for all of them. Cash withdrawal fees. Among the top problems arising with neobanks is that they don’t always have their ATMs. So they partner with banks that provide cash withdrawal but if you use ATM out of the network, this leads to higher fees. Lack of operations with cash. Depending on a neobank, you can get various financial services, from loans to insurance services. However, the truth is that without an external partner, they can’t cover services like cash deposits, available at regular banks. So, if cash operations are crucial to you, be extra attentive if a chosen neobank is partnered with a bank or an ATM network provider. No physical presence. Of course, you can solve most of the issues by contacting a neobank representative online, but for some customers, this can be a trouble. Taking all these into account, a question arises: Is neobank a good choice for a business? Let’s find this out in the next paragraph. Neobanking and Business: What’s the catch? With all pros and cons discussed in the previous chapter, you might think that neobanks are created mostly for individuals. Yet they are good for small and mid-sized businesses. The best thing about them is that a business owner can set all processes and manage all company finances without much problem and papers. In this case, a neobank serves as a platform for finance management that comprises all dashboards, allows for paying for logistics, supply, and ancillary bills. So, choosing a neobank for your business, start with these basic things. Explore the functionality. One neobank is unlike another – this offers you generous fees for money transfers and a referral program, that has a cool dashboard with analytics and a multicurrency account. It can be really hard to choose the one for you, so take your time to know each one available in your region better. Check the fees. Since fees lie at the core of their business model, it’s worth considering how much your potential neobank charges for work. This will help you plan your budget better. Mind the interest rates. Entrusting personal finances to a financial institution, all people want to make sure that they will get as much from this as possible. Make sure that your money will multiply.

Neobanking: Pros, Cons, and How To Use Them Efficiently For Your Business 

HackerNoon Reporter: Please tell us briefly about your  background. My name’s Daumantas Barauskas, I am the COO of Genome. I’ve been a part of the Lithuanian fintech community for over 10 years and now use my expertise in the sphere to grow our EMI. What's your startup called? And in a sentence or two, what does it do? It is called Genome, we are an Electronic Money Institution (EMI) from Vilnius, and started our company in 2018, licensed and supervised by Lithuania’s Central bank. Today Genome is also a principal member of VISA. We offer clients online opening of personal, business, and merchant accounts. Our team also provides relevant business and merchant features: money transfers, currency exchange services, debit card issuing, etc. What is the origin story? Genome’s team consists of many experts in the payments sphere, united by a common goal. That is, to create a product that combines the best features of traditional banks and fintech companies. First, we wanted to significantly expedite the company’s journey to obtaining business and merchant accounts. That’s how we came up with Genome’s unique value proposition – a place where you can manage your finances after passing the onboarding completely online, without time-consuming branch visits. The process is fast and comprehensive, and by the end of it, you are getting access to your accounts, all of which you can manage via the Genome app or in a web browser. And now, a couple of years later, we’ve managed to turn the e-wallet concept into a multi-leveled financial system! Our team continues working on new features for both businesses and individuals. Right now, we are adding virtual and debit business cards to the mix – for our customers’ complete payments experience. What do you love about your team, and why are you the ones to solve this problem? The great thing about the people of Genome is that the company itself was started by fintech specialists with prior experience in the industry. We were well aware of the benefits and drawbacks of banking and went into developing a new product with all these things in mind. As a result, we’ve got an easy-to-use platform with features suitable for people who use financial services not only for their personal needs but also for SMEs and merchants. And we didn’t stop there, as Genome continues to flare with new experts joining the original team. We take the outside perspectives of the newcomers into account and listen to what they have to offer. Together we continue analyzing our clients’ demands and embracing new technologies and trends the industry comes up with. By combining the two approaches, we develop new exciting features. For instance, in recent years, people have taken a liking to a more digital, contactless approach to payments. Thus, we provide our clients with virtual cards, in addition to plastic ones, so they can track each transaction and operation online without any risks of losing their card or other payment data. At the moment, how do you measure success? What are your core metrics? To us, the most vital metrics of success are the clients’ contentment with our services, and the number of people using Genome grows each year. But when it comes to evaluation, we have metrics to keep an eye on. Although Genome is a relatively young company, we have already been acknowledged among other fintech companies in Lithuania. In the second quarter of 2020, Genome showed the second-largest growth in the total amount of payment transactions among Lithuanian payment companies. The number reached over 380 million transactions per year. Last year Genome received the Acquisition International Magazine award as the best online business payments platform 2020 within Europe. Among our recent achievements of 2021 is that Genome was shortlisted for Banking Tech Awards as “Best Digital Initiative” among the solutions by Morgan Stanley, Mastercard, Goldman Sachs Bank USA, and JPMorgan Chase. What’s most exciting about your traction to date? It is difficult to pinpoint a single event, to be honest. You see, it seems like only yesterday we’ve opened Genome’s beta to the public while in reality, it happened in early 2019. Now, three years have passed and surely, they were challenging for us as a startup, not to mention the lockdowns situation. But, I feel that we tried and made the most of the situation at the end of the day. We have consistently been updating Genome, introducing new features, and even introduced two COVID-related initiatives to help the affected businesses by reducing our service fees. All the hard work culminated in the recent release of virtual debit cards. And Genome's physical Visa cards are on the way. We've put a monumental effort into these projects. Right now, virtual cards are available for business users - a great way to pay all the necessary corporate expenses online. All your payment data is securely stored inside your Genome accounts, protected by 2FA. Once our physical debit cards are out, you can use them around the world during business trips. Withdraw money anywhere the ATMs support the Visa cards. The Genome cards are one of the things our team is most excited about for now. What technologies are you currently most excited about, and most worried about? And why? Each year more technologies emerge, but we especially take notice of the ones that intertwine or can benefit the financial sector overall. Right now, our team is on the lookup for the wider implementation of contactless technologies and services. It's the result of people preferring more cautious interactions with other people and institutions. We at Genome followed this principle even before it became essential in 2020, as we provide all our services completely online, meaning you don’t have to visit any branches of ours to open accounts or perform any other operations. For us, it was the question of convenience for clients, and this type of service has become very timely during lockdowns and beyond. Our team continues boosting the contactless approach by offering clients multi-currency virtual cards – a great alternative to plastic with all the features of standard debit cards. And even if you do prefer the traditional physical Visa cards - no worries. After the launch, we will deliver those to you to avoid needless interactions you would have with managers in traditional banks. Now let’s get to the second part of the question. First and foremost, it is not the technology itself that is worrisome, but certain elements that surround it. The elements like data breaches, regulation issues, and keeping the clients’ money secure is what worries us with new technologies emerging. Keep in mind that any technology which deals with people’s data might suffer a breach. And the data leaks, unfortunately, lead to an increase in fraudulent actions that can influence our clients as well. That is why our team always makes sure we have all the necessary security tools and precautions in place, that they are updated and function properly. We abide by strict laws and regulations, provide two-factor authentication for extra security. Genome also offers our business wallet clients a special token that goes with a debit card. This token is a physical device to confirm all the outgoing transactions. It ensures that no unapproved payment will pass, as you need a code generated by the token to complete it. Only a business wallet owner or their authorized colleagues have access to it and know the PIN to receive said code. These are scrupulous efforts, but we can’t go on without them and work each day to put even more security tools and precautions in place. What drew you to get published on HackerNoon? What do you like most about our platform? It's not our first time working with Hacker Noon, we had positive experiences before, writing about fintechs' place within the banking industry. Our team also enjoys other articles on the website. It is a great place if you are interested in modern tech and relevant news. And the separate tech giants section is very helpful too. Vote for Genome on City Award Page! Startup of the Year

On Fintech Trends and Contactless Technology: Startup Interview with Daumantas Barauskas

This story is a part of Hacker Noon's series of interviews. The series is intended for tech professionals contributing the most insightful Hacker Noon stories to share more about their writing habits, ideas, and professional background (and maybe a hobby or two). If you too would like to start contributing to Hacker Noon, you can do so . Meet the Writer here So let’s tell us a bit about yourself. For example, name, profession, and personal interests. Hello, dear Hackernoon readers. My name is kinda weird Slavic one. Ladislav. Weird in a sense that neither I nor any fellow Slavic has any idea what the hell short form of Ladislav is. As for interests. I love anything creative. My biggest success perhaps is my behavioral zero-day exploit blocker that protects hundreds of millions of computers as we speak ever since ESET nod antivirus v 5. I wrote the patent worked hard to implement and release it safely to the world on millions of computers without bluescreens. Yet my patent was stolen by the company while I was nearly dead in hospital after a paragliding accident. Yes… East Europe is definitely not a Silicon Valley. Yet. I don’t believe that there is such a thing as bad people. Just bad software. I.e way too many generations had been raised in an environment without moral examples.  But recently, luckily, Covid made me unemployed, and shackles of my creativity were (unfortunately for you, my dear readers ;D) finally released. Yay. In a sense, this Covid period opened my eyes. Now I finally understand why so much of math or philosophy was discovered in ancient Greece. Creative new ideas are now are swirling in my head pretty much every day. Ideas that I had never time to think about due to exhaustive daily jobs occupying my brain potential on boring and repetitive tasks. Don’t you ever wonder why is it that most of mathematics physics philosophy cornerstones were invented in this ancient but relatively short part of our history? Followed by pretty much just a few blips here and there?. I believe that the key to this was to allow extremely large population freedom to occupy their brain with other things than just delivering food and shelter. The majority of Ancient Greece was essentially a bunch of unemployed folks contemplating the universe’s inner workings and such since someone else did the boring repetitive work for them. Why I am mentioning this.? I feel like humanity’s second renaissance is coming. Why? Well, every time a major event like Covid or AI and automation taking the majority of repetitive boring jobs happens. The universe splits into many but usually just two major branches depending on how clear-cut the nearest probabilities for other outcomes were. I suppose in 50/50 like cases Kazimir effect tips the scales causing the butterfly effect. So. Let’s pretend that we will continue for a bit in this optimistic fork of this universe. A fork where we soon will have AI and robots doing most of our manual, repetitive work, with us being all on universal income finally using our brains in more creative ways like in ancient Greece. Covid, in a sense, accelerated this transition with another payment from the government coming soon.  And as for the less optimistic branch of this universe’s fork. Luckily we have Elon for scaring the doors from us. Interesting! What was your latest Hackernoon Top story about? My last story was a crazy idea that I got one night and was super excited about, and immediately messaged to like half of SpaceX and asked them what they think about it. As . (So now I have like 10 SpaceX connections that can’t talk to me due to either NSA orders or the idea being too stupid). With both being equally probable. you do But In short. It just baffled me why we still keep building thin 130m tall and unstable rockets and then wait for good weather or government approval. So I proposed a starship transformer. The ship has the same volume but height of just 10m during start or landing. It, therefore, does not require any special towers not is it any longer fixed to time(weather) or place. Essentially, It could transform to a tall falcon-heavy-like upright supersonic configuration on demand. Do you usually write on similar topics? If not, what do you usually write about? It’s hard to say since I am fascinated by so many things. Take, for example, our recent rapid advancements in the u of every living thing around us. And gene editing with Omega in particular. Did you know that 70% of mosquitos’ DNA is the same as yours and it is not a coincidence? Thing is. Life is pretty much a bunch of so far 3 types of programable (cells) with mostly the same code over and over. nderstanding of the genome nanobots So right now, I am writing a new story about how radically we can change the world by flipping a few homeobox-like gene bits in the DNA of Bamboo to control its size, shape, and direction. Why Bamboo? Well, Bamboo has strength and weight rivaling carbon fiber and, more importantly, can grow 1m a day even without tweaks. Unfortunately still has hardwired dimensions. ;D So with that little DNA tweak, you can essentially not build but grow green-sustainable, recyclable houses skyscrapers, and due to the parallel nature, even whole cities in a week. Branches solve traffic congestion by moving in the dense 3D mazes without the high energy cost of flying, and energy collection is already built-in in a pretty large surface of ever-present  leaves. Engineer nets will grow even thru the strongest granite via an unstoppable army of strong roots. And all of that with just a flip of a few bits. Great! What is your usual writing routine like (if you have one?) For me, I usually get super excited about something. Then it pretty much goes downhill with this weird drive of mine to share emotions and excitement and such. You know … Of the magic moment. And. If the is not at that particular moment at the wheel, then I start immediately to write everything down. I keep adding a lot of ideas in pure, unfiltered flow. You never know what branches lead where. Then I filter them thru good old google to see  how much of a wheel I am just reinventing. procrastination monkey In reality, a lot of these secondary ideas are usually either similar ideas with different wording or just way too much noise about nothing. So finally, my favorite Occam's razor jumps in, helping to distill just the best parts in the most succinct way possible since people rarely read past 3rd minute. And since English is not my first language In the final phase, I try to do my best to hide it ;) Being a writer in tech can be a challenge. It’s not often our main role, but an addition to another one. What is the biggest challenge you have when it comes to writing? According to famous Professor and Neuroscientist at Stanford University Andrew Huberman, . The creativity of our brains is essentially a bunch of hamsters on wheels. And they pretty much run on dopamine.  Rare stuff that is naturally produced when you are in an equally rare state of happiness. Especially rare if you are stuck in east Europe. Luckily there are a few other ways to get it. I don’t do drugs obviously. Then there is sex. But that is out of the question since I am butt ugly. Still, I wish to have children one day and watch them grow. I guess my next respawn in this simulation. It’s quite simple actually So what is left for me is pretty much dopamine cheap offshoot adrenaline. Stuff I get from motorcycles paragliding or being broke and not filing tax reports in time. Endorphins I get from exercise or running. Or if you are not spoiled, a jam or Nutella will do wonders too and lasts ages. Thankfully I am getting crazy new ideas once a week, which produces a day or two of excitement and endorphins too. What is the next thing you hope to achieve in your career? Exploring  crazy and fun ideas and hoping one day to implement them is fun and all. But what I really hope for right now. Is to find someone who will help finish the reusable particle-based ; Without filters or advice, I am stuck. Considering what I am trying to do and that so far, even a panel of respected judges and virologists didn’t find a major reason why it wouldn’t work; it still baffles me that nobody wishes to help. covid test I passed to the semifinals in XPrize It’s times like these when I am starting to believe I am in some kind of weird simulation. I mean, the cost of filters I need is out of my reach yet it is way less than someone pays for breakfast daily.  The optimist in me says that perhaps I am just struggling to get a word around. Why I am so focused on that covid test so much now over a year?. To be frank, considering the state we are in, not many things make more sense to me as having higher priority. Wow, that’s admirable. Now, something more casual: What is your guilty pleasure of choice? No Idea. I have always been poor. I always dreamed about traveling the world though. Or have money to start building my own variant of and fly. And by flying, I mean really fly. Like with moving powerful hydraulics copying every slightest move or even hint of your  arms or hands. So you can clearly see now that my biggest guilty pleasure is that with my brain being free again I started dreaming again ;) 12m wide foldable hydraulic wings Do you have a non-tech-related hobby? If yes, what is it? I am fanatically attracted to the ancient history of and life in general. I traveled a bit too. Then I am way too often attracted to new papers in fields I am not supposed to. I know. I know. Hereby dragons.  Not that I understand a thing. But curiosity knows no boundaries. Like the very recent fascinating new OMEGA gene editing advancements. And despite not being able to absorb or grasp all their content. There is still a lot of fancy pictures too. A lot of fun. DNA If I oversimplify it a bit. The scientists are, in a sense, writers too and are usually preparing their papers for the most efficient consumption. So the well-written paper usually has a very short but punctual condensed intro, and conclusion sections with plenty of fancy self-explanatory charts sparkled throughout. Yes, you need to possess at least a basic engineering decoding toolkit for something to stick. And something what Stanford psychology professor Carol Dweck calls a growth mindset--The belief that you can grow and develop skills and abilities you didn't have before. But in my humble opinion, in this age, where every major educational information is public and found in seconds. An engineer is perhaps no longer only defined by which prestigious school he had taken but by his mindset. I.e., how he searches, processes, filters, and absorbs and combines that information with what he already knows. Especially how often he seeks and absorbs new information. And there are so many fascinating things happening as we speak in biology, genetics, computer graphics, or info. neural networks covid related What can the Hacker Noon community expect to read from you next? I am in the middle of a Bamboo software hacking article. To search for this elusive homeobox gene location, I currently have a lot of fun with searching DNA on GPU via 10 lines of Webgpu. Why Webgpu? First. Webgpu is HOT. I mean common. WebGL is so yesterday. Second. So article readers who are on HN, mostly programmers, will be able to do it themself in a browser too. And let’s face it. Which programmer would resist being able to massively accelerate their code in the 10 lines? The world of biology and genetics (or the world of nanobots as I call it) is so exciting. So much fun. Thanks for taking the time to join our “Meet the writer” series. It was a pleasure. Do you have any closing words? Thanks for stopping by. And those of you who wish to with the reusable test, especially before spring starts. 80nm-120nm Filters and Filter sealing design advice would be helpful. I am also pretty sure my design has a lot of flaws. But that is not the point. Nothing exists in v 1.0. We need to keep working on hourly reusable decentralized cheap test iteration after iteration. Since vaccinating the whole world is currently financially, politically, and logistically pretty much impossible. The More brains and hands-on-deck, the faster the progress. Let’s find a way to stop Covid together. join forces and help with stopping this covid nightmare

Meet the Writer: Hacker Noon's Contributor Ladislav Nevery, Innovator

Has the time of traditional banks passed? Some people may think that’s the case, referring to the steady popularity growth of the neobanks – a type of financial institution that operates online, having no physical branches. Just in Europe alone, the penetration of online banking among the EU markets amounted to 58% in 2019. And in Norway, Iceland, Denmark, Finland, and the Netherlands the number passed the 90% mark. [ ] 1 It won’t be a stretch to assume that in 2020 even more people turned to digital banks because of convenience and/or safety reasons. The examples are many: the lockdown prompted British citizens to use banking apps, with 54% of people surveyed in July said it’s a regular thing for them now. [ ] 2 Meanwhile, in April stimulus payments helped increase the mobile banking registrations by 200% over the daily average a month prior. [ ] 3 All in all, the size of a global online banking market is expected to more than double: from $9,1 billion in 2019 to $20,5 billion by 2026, as the demand for banking services among younger customers, who are more up to speed with the latest technological advances, will grow. [ ] 4 The reasons people are choosing neobanks over the brick-and-mortar ones are pretty common. For instance, the 2017 survey conducted in Italy, and the 2020 British survey pinpoint the few main justifications for using digital banking: the convenience of online services, better rates, and economic conditions. The seamlessness of digital banking services is what they are famous for in the first place – for instance, using Genome, a person or a business can manage their finances online, from the convenience of their houses and offices. You don’t need to waste time visiting banking branches, which in 2020 has also become a matter of safety. As for more beneficial rates, traditional banks usually charge more for their services, while Fintech companies offer higher interest rates and bonuses to attract more new clients. [ , ] 5 6 Still, all the factors listed are not enough to erase the impact traditionally banks still have on the financial industry. Remember, lots of brick-and-mortar have decades or even hundreds of years of development behind them, making them a lot more reputable and trustworthy, than newly emerging Fintechs. They also offer clients loans and mortgages and can work with businesses of different sizes, which most of the neobanks can’t afford. The capital assets are another crucial argument in traditional banks’ success. As data shows, right now the Industrial and Commercial Bank of China (ICBC) is the largest bank by assets, which amount to over $4.3 trillion. In a matter of fact, every bank from the top 10 of the said list has over $1,95 trillion in assets. [ ] 7 The European banks don’t fall behind either: compared by total assets, the UK’s HSBC Holdings (€2,4 trillion), French BNP Paribas (€2,1 trillion), French Credit Agricole (€2 trillion), Spanish Banco Santander (€1,5 trillion), and French Societe Generale (€1,3 trillion) are in a top 5. [ ] 8 If compared by market capitalization, the ICBC reported $1714 billion in the second quarter of 2020. Among other banks like JPMorgan Chase and Bank of America, the index reached $287 billion and $206 billion accordingly in the second quarter of the year. Consider, that the pandemic hit the traditional bank industry hard, as in the fourth quarter of 2019 the market capitalization index of three said banks amounted to $2051 billion, $437 billion, and $317 billion accordingly. [ , , ] 9 10 11 Though big traditional banks withstood the first wave of the pandemic, it uncovered the many problems they have. Not all brick-and-mortar banks were digitally ready when the lockdowns started: many banking branches closed, leaving customers unable to reach their banking manager. As a result, people resorted to using e-banking apps, which were already perfect for a lockdown scenario – no need for branches’ visiting and faster operation performance than with traditional banks, all is required is to have a phone. Some of the banks got to correct their mistakes: the survey of banks from around 39 countries revealed that 34% of respondents had to implement fully digital processes due to the pandemic, 23% enabled digital methods for identification and verification, and 18% even launched contactless payment methods. [ ] 12 The need for digital services is a valid reason for people to switch from traditional to neobanks, especially in 2020, but all in all, people still prefer the good ol' banks for good customer service and an opportunity to see the manager in person. [ ] 6 So, how do you solve the issue? It seems that taking the best things from both worlds of traditional and e-banks is the best-case scenario, which can lead to cooperation between the two. Brick-and-mortar banks can use their ample resources to invest in Fintech company, which in turn will provide the latest solutions for digital banking. Following the innovations, while keeping up with the high-level customer support and remaining secure and trustworthy, is what traditional banks should gravitate to. Also, as the 2019 research from Accenture uncovered, only 12% of analyzed banks were committed to a digital transformation, 38% sought digitalization, but their efforts weren’t as evident, and a staggering 50% didn’t make much progress at all. [ ] 13 That’s why, looking up to neobanks’ successful experiences from the technical point of view is something that will, hopefully, push the very needed change among old-fashioned banks, even if they aren’t considering cooperation with Fintechs. For example, is licensed by the National Bank of Lithuania to maintain that level of reliability with the technological advances of a Fintech startup. Genome

Neobanks Notwithstanding, Traditional Banks Are Not Going Away

Gene sequencing , but according to a report by Grand View Research (published in September 2020) the global DNA sequencing market size was in 2019 and is expected to expand at a compound annual growth rate (CAGR) of 11.4% from 2020 to 2027. began in the 1970s valued at USD 4.7 billion Rapid advancements in sequencing technology and bioinformatics have enabled the identification of DNA variations -- and popular sites like 23andMe and Ancestry.com have continued to grow -- . for $99 you can get a profile of your DNA But experts assert that these tests may be "loss leaders" for these sites -- it costs more to do the tests than they are charging users. But how do these sites make money? Though both popular sites claim they do not sell their data, there is more ambiguity about who has access to the database themselves, and who has rights to the anonymized data, which might be valuable to pharmaceutical companies,  insurance companies, individuals involved in paternity or inheritance disputes, and law enforcement agencies. 23andMe states that it does not grant access to the police, but other companies, such as FamilyTreeDNA, state that they do. There are examples of DNA data collected on these sites being put to good use like who was identified as California’s notorious and arrested in 2018 after investigators matched a crime scene DNA sample to the results of DTC testing uploaded on to a public genealogy site by the man's relative. Joseph James DeAngelo Golden State Killer Government-run biobanks have also granted access to police, which also led to the conviction of Swedish foreign minister ’s assassin in 2004. Anna Lindh Experts speculate that biological data could be used for identifying terrorist suspects, tracking military personnel, and the rationing of treatment in overstretched health systems. We interviewed who has degrees in Philosophy, Law, and Science. He has written extensively on data ownership, including biometrics, and in particular on gene ownership. In 2009, he published a book titled " " which was cited in a high-profile lawsuit taken by the American Civil Liberties Union against a company that had patented genes. Dr. David Koepsell Who Owns You? The arguments proved successful and were eventually adopted by the US Supreme Court in the overturning of the practice of gene patenting. He is also CEO of DNA data marketplace . EncrypGen “ ,” Dr. Koepsell said. Since the first mapping, we’ve been zooming in, getting a better understanding of the different genes, trying to make sense of the complicated code Koepsell went on to explain the reason these sites offer such competitively priced tests can often be seen in the small print. Site users who purchased genetic tests are also asked if they would allow their DNA data to be used for scientific research. In most cases, people ticked yes -- it seemed a fair enough request. However, it later turned out this data, packaged in the right way, was many times more valuable even than the initial fee. “ ," Dr. Koepsell said. The intention is for pharma companies to make profitable drugs, and that is good for everyone, but I also think that people whose data is being used ought to be paid for the value of that data. It all goes back to transparency These companies act essentially as DNA data brokers to the pharmaceutical industry, offering the data to the highest bidders. The price differential was significant and while this industry is largely unregulated and hard to measure, DNA data has achieved hundreds of millions of dollars in sales. “ ," Koepsell said. Most people who have availed of the test agree to the scientific research; I didn’t and, as a result, I get frequent emails asking me to change my mind. I’m not going to, even to help their profit or especially not to help their profit in 2018 for $300 million to have exclusive access to their datasets. As a trend, it seems the continued use of DNA data that was shared in good faith and with a vague understanding of the implications will be used in a number of industries in violation of most users' wishes and intentions. GlaxoSmithKline entered into a DNA data deal with 23andMe A spokesperson for Ancestry.com contacted us after the publication of this story to say: In their , Ancestry.com states: " We do not voluntarily share your information with law enforcement. ...we will not share your Genetic Information with insurance companies, employers, or third-party marketers without your express consent." "Ancestry does not sell consumer DNA data in any form... Our revenue comes solely from the sale of our products and services to consumers. Ancestry has provided customers with the voluntary option to participate in a small number of academic and scientific research studies to contribute to scientific discoveries – as publicly disclosed on our website." privacy statement Ancestry does not share your individual Personal Information (including your Genetic Information) with third-parties except as described in this Privacy Statement or with your additional consent. According to the Ancestry.com privacy statement, although individualized genetic information is not used for other purposes: "Studying aggregated Genetic Information to better understand population and ethnicity-related health, wellness, aging, or physical conditions," though they state this is only done with user consent. This is an important distinction for users who pay close attention to what they are consenting to when they give their genetic information to sites like Ancestry.com. Conclusion I had followed the story of the for just a couple of years before he was caught. It was a great relief that he was finally brought to justice. Golden State Killer Unfortunately, that doesn't really detract from the underlying message -- when there is sufficient incentive, the government has a critical mass of DNA data from these websites which are capable of finding people at least some of the time. It's a public good to use this data to bring more criminals to justice. One wonders what abuses of that power are inevitable if the trajectory of increasing DNA information and no oversight on how it is used continues. Image by Pete Linforth from Pixabay

23andMe and Other Sites are Selling Users' Genetic Data: How Safe is Your DNA?

My friend Anshul recently told me about . I had heard about it clogging up the ETH pipes and causing the price to drop but had not realized what an amazing design the CryptoKitties team had realized. CryptoKitties are basically breed-able beanie babies built on top of the ethereum . And yet, they are also so much more than that: CryptoKitties are a digital playground where programmers, geneticists, and experts in economics, bioinformatics, machine learning and game theory can come together to learn and play. CryptoKitties blockchain Studying cryptokitties For those of you new to CryptoKitties, have a look at the ; this will teach you the basics of the CryptoKitties game system. To get a better view of exactly just what a kitty is you can go through the . You’ll find that a CryptoKitty is basically a struct on the EVM that looks like this. Mashable guide CryptokittiesCore contract struct Kitty {uint256 genes;uint64 birthTime;uint64 cooldownEndBlock;uint32 matronId;uint32 sireId;uint32 siringWithId;uint16 cooldownIndex;uint16 generation;} We see some other interesting things too… modifier onlyCEO() {require(msg.sender == ceoAddress);_;} and… /// SEKRETOOOOcontract GeneScienceInterface {function isGeneScience() public pure returns (bool); @title /// given genes of kitten 1 & 2, return a genetic combination - may have a random factor/// genes1 genes of mom/// genes2 genes of sire/// the genes that are supposed to be passed down the childfunction mixGenes(uint256 genes1, uint256 genes2, uint256 targetBlock) public returns (uint256);} @dev @param @param @return Wow a secret. This is THE secret. If you’ve been playing cryptokitties you know this is how kitties are produced. You’ll notice that siring kitties produces offspring in a manner that feels intuitive: features get passed on from parents to offspring sometimes but not always, sometimes skipping generations. This is the code that runs at the conception of cryptokitties with non-zero generations. A brief genetics remedial from your 10th grade biology class A (from , meaning ‘to show ‘, and , meaning ‘type’) is the composite of an ’s observable characteristics or , such as its , , biochemical or physiological properties, , and products of behavior (such as a bird’s nest). A phenotype results from the of an organism’s genetic code, its , as well as the influence of environmental factors and the interactions between the two. When two or more clearly different phenotypes exist in the same population of a species, the species is called . A well-documented polymorphism is ; while the coat color depends on many genes, it is clearly seen in the environment as yellow, black and brown. phenotype Greek phainein typos organism traits morphology development behavior expression genotype polymorphic Labrador Retriever coloring Managing all the features in my head was difficult. I needed somewhere to organize my thoughts so I created a do do some of the math. From the currently features that are being served up by the crypokitties.co API we know that a CryptoKitty has up to 1.15792×10⁷⁷ phenotypes. Given that the ‘genotype’ for a CryptoKitty is a 256-bit integer there are 2²⁵⁶ possible genotypes for a CryptoKitty. Going to cryptokittydex I was able to see the genotypes for each kitty and block addresses! What fun! spreadsheet totesbasic The API at CryptoKitties.co serves the public data needed to create the game around this blockchain. It creates the images, names the “features” serves then via the API and facilitates the ethereum transactions required to play the game using MetaMask. Inspecting the browser while playing cryptokitties, you’ll find that the requests to cryptokitties.co are serving results related to the features expressed in the phenotype and familial relationships. Running curl ' https://api.cryptokitties.co/kitties/123606' gives {"id": 123606,"name": null,"generation": 4,"created_at": "2017-12-07T06:54:46.000Z","image_url": " ","color": "topaz","bio": "What's up! I'm Kitty #123606. I've never told anyone this, but I once pawed at a dog. I once got in a fight with a moose, and won. We can be friends, but keep the ultra purrsonal stuff to yourself, please.","is_fancy": false,"is_exclusive": false,"fancy_type": null,"status": {"is_ready": true,"is_gestating": false,"cooldown": 1512636431155,"cooldown_index": 3},"auction": {},"owner": {"address": "0xb464f93be01bde143c92176d921c1d0d533bc3ca","image": "17","nickname": "KittyHodlr"},"matron": {"id": 123288,"name": null,"generation": 3,"owner_wallet_address": "0x8081ecb6f3350fe5ed46f1bd74d5ada62c62e740","created_at": "2017-12-07T06:30:08.000Z","image_url": " ","color": "topaz","is_fancy": false,"is_exclusive": false,"fancy_type": null,"status": {"is_ready": true,"is_gestating": false,"cooldown": 1512635143445}},"sire": {"id": 96883,"name": "Gen 3 Purple Sizzurp","generation": 3,"owner_wallet_address": "0x8081ecb6f3350fe5ed46f1bd74d5ada62c62e740","created_at": "2017-12-05T23:09:11.000Z","image_url": " ","color": "sizzurp","is_fancy": false,"is_exclusive": false,"fancy_type": null,"status": {"is_ready": false,"is_gestating": false,"cooldown": 1512641108913}},"children": [],"cattributes": [{"description": "orangesoda","type": null}, {"description": "simple","type": null}, {"description": "granitegrey","type": null}, {"description": "coffee","type": null}, {"description": "pouty","type": null}, {"description": "totesbasic","type": null}, {"description": "topaz","type": null}, {"description": "cymric","type": null}]} https://storage.googleapis.com/ck-kitty-image/0x06012c8cf97bead5deae237070f9587f8e7a266d/123606.svg https://storage.googleapis.com/ck-kitty-image/0x06012c8cf97bead5deae237070f9587f8e7a266d/123288.svg https://storage.googleapis.com/ck-kitty-image/0x06012c8cf97bead5deae237070f9587f8e7a266d/96883.svg This is what a CryptoKitty looks like according to the API. We can see the owners address, and the features of Kitty # . Most importantly we get an image of an amazing, fluffly cat. 123606 So cute and fluffly! Each cat has cattributes, based on its genes. Remember that 256 bit uint ? That determines what your kitty looks like. Ignoring the auction system, we can think about the genetic aspect of the system as having two parts genes X **— (cryptokitties.co)→ (**phenotype X, image X) genes matron+ genes sire genes offspring —( mixGenes )→ where is our secret sauce function. Could we study the reproductive patterns of cryptokitties to be able to predict breeding outcomes better than chance? mixGenes We want to be able to get data about the genomes of these mystical creatures by inspecting the ethereum blockchain. We could write code to download and parse all cryptokitties transactions so far but that would be slow and ain’t nobody got time for cat. I decided to quickly write a scraper to collect the current set of cats available on cryptokitties.co and cryptokittydex.com. I wrote this to use a headless Chrome instance to ‘visit’ and put the genomes of all the cats, their owners and their images in an S3 bucket (or on disk). I let it run till it downloaded about 100k cats and their images. code cryptokittydex.com The next day I decided to study the sample of cryptokitties I had scraped. I quickly noticed that fancy kittys had extensions while normal kitties were saved as I began to notice patterns and learnt the basics of each cryptokitty attribute by trial and error and referring to reddit guides. It became obvious which features were exclusive and I slowly felt I was becoming better at categorizing cryptokitties. By the end of the night I felt I could soon be the Gregor Mendel of cryptokitties. .png .svg. Cryptokitty basics There are 3 basic feature types. Excluding ‘fancy’, there are 7 or 8 visible features for each cat. There are 2 reproductive features for each cat that are not visual. : BodyType, PatternType, MouthType, EyeType, Body Features (4) : Primary, Secondary, Tertiary, Eye Colors Features (4) : CoolDown, Generation Reproductive Some patterns I noticed while going through the cats. Totesbasic cats do not express their tertiary color but do have the gene The beard is the only mouth feature that contains color. It uses the primary color. The eye color is also the color of the background The primary color is the main color of the cats fur The secondary color is the color of the cats fur patch(s) / stripes if their pattern type has 2 colors Tertiary colors are expressed in the cats fur for some pattern types with 3 colors Now the question was — which kitties had not been discovered yet? And could we somehow do a to see what these kitties would look like before they are born? Opening up the SVGs it was easy to see the different parts of the kitty. cryptokitty genetic test Grepping around I looked for the color in this kitty and searched for it in another. It was the same across all the files! However continuing through the dataset I found not all fur patterns used all 3 colors and some were pretty convoluted making it difficult to manipulate the svgs directly with JSDom. I decided to just treat the svgs as strings and create three sets of images. The first would be a set of 48 images that would have cats of every possible BodyType and PatternType combination. The second was a set of 8 cats with all Mouth combinations. The third — you guessed it — the set of all possible eyes. Isolating the svg elements for each of these ‘genes’ allowed me build a quick React.js prototype that was able to render a kitty of any possible phenotype. aquamarine ={body} ={mouth} ={eye} ={pattern} ={[c.Primary[primary],c.Secondary[secondary],c.Tertiary[tertiary],c.EyeColor[eyeColor]]}/> <Cryptokitty body mouth eye pattern colors Adding some simple controls, I had a very early prototype for my cryptokitty genetic testing company. Meow all I have to do is find a cool person with some good venchair carpetal. Tell me what you think! PRs are highly encouraged. CryptoKitty Designer: https://bytesforbites.github.io/cryptokitty-designer/#/ Source: . https://github.com/BytesForBites/cryptokitty-designer Send kittehs and tips to 0x02Ee97a13e434717e3FFa12758a235D1a1680775 Please sir, I’m just a dev cat that needs to pay for my new mouse

Hacking the CryptoKitties Genome

Table of Links Abstract and 1 Introduction 1.1 Related literature Types of Metrics and Their Hypothesis and 2.1 Types of metrics
2.2 Hypotheses for different types of metrics


Type I and Type II Error Rates for Decision Rules including Superiority and Non-Inferiority Tests
3.1 The composite hypotheses of superiority and non-inferiority tests
3.2 Bounding the type I and type II error rates for UI and IU testing
3.3 Bounding the error rates for a decision rule including both success and guardrail metrics
3.4 Power corrections for non-inferiority testing


Extending the Decision Rule with Deterioration and Quality Metrics


Monte Carlo Simulation Study
5.1 Results


Discussion and Conclusions APPENDIX A: IMPROVING THE EFFICIENCY OF PROPOSITION 4.1 WITH ADDITIONAL ASSUMPTIONS APPENDIX B: EXAMPLES OF GLOBAL FALSE AND TRUE POSITIVE RATES APPENDIX C: A NOTE ON SEQUENTIAL TESTING OF DETERIORATION APPENDIX D: USING NYHOLT’S METHOD OF EFFICIENT NUMBER OF INDEPENDENT TESTS Acknowledgments and References APPENDIX D: USING NYHOLT’S METHOD OF EFFICIENT NUMBER OF INDEPENDENT TESTS As discussed above, the lack of assumption regarding the data generating process implies that under many processes the decision rule with have too low type I error rate and to high power. Although there are many alternatives to Bonferronitype corrections we have two conditions that we would like to have for any design. First, the individual metric results should have CI’s that maps to the decision rule outcome. Second, the design should allow for analytical required sample size calculations. These conditions means that most alternative multiple correction methods, such as [7] and [8] are discarded. Very few methods have confidence intervals, and even fewer are possible to integrate in the sample size calculation, due to their reliance on the p-values. One method that fulfills our conditions is Nyholt’s method for calculating the so-called effective number of independent tests in a set of tests with arbitrary covariance structure [14]. This method is frequently applied in high-dimensional genome testing, and has been refined by several, see e.g. [11, 5]. Nyholt’s method is simple, the effective number of independent tests is given by D.0.1 Monte Carlo simulation results with the Nyholt method Below the tables from Section 5 are repeated together with the result using Nyholts method as described above. As expected, the method improves the efficiency. This is shown by type I error rates and power closer to the intended rates under certain covariance structures. ACKNOWLEDGMENTS The authors would like to thank Bob Wilson for feedback on earlier drafts of this paper. REFERENCES [1] BERGER, J. (2013). Statistical decision theory: foundations, concepts, and methods. Springer Science & Business Media. [2] DMITRIENKO, A., OFFEN, W. W. and WESTFALL, P. H. (2003). Gatekeeping strategies for clinical trials that do not require all primary effects to be significant. Statistics in Medicine 22 2387-2400. [3] DMITRIENKO, A., TAMHANE, A. C. and BRETZ, F. (2009). Multiple testing problems in pharmaceutical statistics. CRC press. [4] FABIJAN, A., GUPCHUP, J., GUPTA, S., OMHOVER, J., QIN, W., VERMEER, L. and DMITRIEV, P. (2019). Diagnosing Sample Ratio Mismatch in Online Controlled Experiments: A Taxonomy and Rules of Thumb for Practitioners. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. KDD ’19 2156–2164. Association for Computing Machinery, New York, NY, USA. https://doi.org/10.1145/3292500.3330722 [5] GALWEY, N. W. (2009). A new measure of the effective number of tests, a practical tool for comparing families of non-independent significance tests. Genetic Epidemiology: The Official Publication of the International Genetic Epidemiology Society 33 559–568. [6] GUILBAUD, O. (2014). Sharper Confidence Intervals for Hochberg-and Hommel-Related Multiple Tests Based On an Extended Simes Inequality. Statistics in Biopharmaceutical Research 6 123–136. [7] HOLM, S. (1979). A simple sequentially rejective multiple test procedure. Scandinavian journal of statistics 65–70. [8] HOMMEL, G. (1988). A stagewise rejective multiple test procedure based on a modified Bonferroni test. Biometrika 75 383–386. [9] KONG, L., KOHBERGER, R. C. and KOCH, G. G. (2004). Type I Error and Power in Noninferiority/Equivalence Trials with Correlated Multiple Endpoints: An Example from Vaccine Development Trials. Journal of Biopharmaceutical Statistics 14 893-907. PMID: 15587971. https://doi.org/10.1081/BIP-200035454 [10] LEHMACHER, W., WASSMER, G. and REITMEIR, P. (1991). Procedures for Two-Sample Comparisons with Multiple Endpoints Controlling the Experimentwise Error Rate. Biometrics 47 511–521. [11] LI, J. and JI, L. (2005). Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix. Heredity 95 221–227. [12] LIU, S. and LIU, A. (2022). The Lifecycle of Developing Overall Evaluation Criterion in AB Testing for Netflix Messaging. In Proceedings of the 26th International Conference on Evaluation and Assessment in Software Engineering 266–267. [13] NEUHÄUSER, M. (2006). How to deal with multiple endpoints in clinical trials. Fundamental & Clinical Pharmacology 20 515-523. [14] NYHOLT, D. R. (2004). A simple correction for multiple testing for single-nucleotide polymorphisms in linkage disequilibrium with each other. The American Journal of Human Genetics 74 765–769. [15] O’BRIEN, P. C. (1984). Procedures for Comparing Samples with Multiple Endpoints. Biometrics 40 1079–1087. [16] OFFEN, W., CHUANG-STEIN, C., DMITRIENKO, A., LITTMAN, G., MACA, J., MEYERSON, L., MUIRHEAD, R., STRYSZAK, P., BADDY, A., CHEN, K., COPLEY-MERRIMAN, K., DERE, W., GIVENS, S., HALL, D., HENRY, D., JACKSON, J. D., KRISHEN, A., LIU, T., RYDER, S., SANKOH, A. J., WANG, J. and YEH, C.-H. (2007). Multiple Co-primary Endpoints: Medical and Statistical Solutions: A Report from the Multiple Endpoints Expert Team of the Pharmaceutical Research and Manufacturers of America. Drug Information Journal 41 31–46. https://doi.org/10.1177/009286150704100105 Authors:
(1) Mårten Schultzberg, Experimentation Platform team, Spotify, Stockholm, Sweden;
(2) Sebastian Ankargren, Experimentation Platform team, Spotify, Stockholm, Sweden;
(3) Mattias Frånberg, Experimentation Platform team, Spotify, Stockholm, Sweden. This paper is available on arxiv under CC BY 4.0 DEED license. Table of Links Abstract and 1 Introduction Abstract and 1 Introduction 1.1 Related literature 1.1 Related literature Types of Metrics and Their Hypothesis and 2.1 Types of metrics
2.2 Hypotheses for different types of metrics Type I and Type II Error Rates for Decision Rules including Superiority and Non-Inferiority Tests
3.1 The composite hypotheses of superiority and non-inferiority tests
3.2 Bounding the type I and type II error rates for UI and IU testing
3.3 Bounding the error rates for a decision rule including both success and guardrail metrics
3.4 Power corrections for non-inferiority testing Extending the Decision Rule with Deterioration and Quality Metrics Monte Carlo Simulation Study
5.1 Results Discussion and Conclusions Types of Metrics and Their Hypothesis and 2.1 Types of metrics 2.2 Hypotheses for different types of metrics Types of Metrics and Their Hypothesis and 2.1 Types of metrics Types of Metrics and Their Hypothesis and 2.1 Types of metrics 2.2 Hypotheses for different types of metrics 2.2 Hypotheses for different types of metrics Type I and Type II Error Rates for Decision Rules including Superiority and Non-Inferiority Tests 3.1 The composite hypotheses of superiority and non-inferiority tests 3.2 Bounding the type I and type II error rates for UI and IU testing 3.3 Bounding the error rates for a decision rule including both success and guardrail metrics 3.4 Power corrections for non-inferiority testing Type I and Type II Error Rates for Decision Rules including Superiority and Non-Inferiority Tests Type I and Type II Error Rates for Decision Rules including Superiority and Non-Inferiority Tests 3.1 The composite hypotheses of superiority and non-inferiority tests 3.1 The composite hypotheses of superiority and non-inferiority tests 3.2 Bounding the type I and type II error rates for UI and IU testing 3.2 Bounding the type I and type II error rates for UI and IU testing 3.3 Bounding the error rates for a decision rule including both success and guardrail metrics 3.3 Bounding the error rates for a decision rule including both success and guardrail metrics 3.4 Power corrections for non-inferiority testing 3.4 Power corrections for non-inferiority testing Extending the Decision Rule with Deterioration and Quality Metrics Extending the Decision Rule with Deterioration and Quality Metrics Extending the Decision Rule with Deterioration and Quality Metrics Monte Carlo Simulation Study 5.1 Results Monte Carlo Simulation Study Monte Carlo Simulation Study 5.1 Results 5.1 Results Discussion and Conclusions Discussion and Conclusions Discussion and Conclusions APPENDIX A: IMPROVING THE EFFICIENCY OF PROPOSITION 4.1 WITH ADDITIONAL ASSUMPTIONS APPENDIX A: APPENDIX A: IMPROVING THE EFFICIENCY OF PROPOSITION 4.1 WITH ADDITIONAL ASSUMPTIONS APPENDIX B: EXAMPLES OF GLOBAL FALSE AND TRUE POSITIVE RATES APPENDIX B: APPENDIX B: EXAMPLES OF GLOBAL FALSE AND TRUE POSITIVE RATES APPENDIX C: A NOTE ON SEQUENTIAL TESTING OF DETERIORATION APPENDIX C: APPENDIX C: A NOTE ON SEQUENTIAL TESTING OF DETERIORATION APPENDIX D: USING NYHOLT’S METHOD OF EFFICIENT NUMBER OF INDEPENDENT TESTS APPENDIX D: APPENDIX D: USING NYHOLT’S METHOD OF EFFICIENT NUMBER OF INDEPENDENT TESTS Acknowledgments and References Acknowledgments and References APPENDIX D: USING NYHOLT’S METHOD OF EFFICIENT NUMBER OF INDEPENDENT TESTS As discussed above, the lack of assumption regarding the data generating process implies that under many processes the decision rule with have too low type I error rate and to high power. Although there are many alternatives to Bonferronitype corrections we have two conditions that we would like to have for any design. First, the individual metric results should have CI’s that maps to the decision rule outcome. Second, the design should allow for analytical required sample size calculations. These conditions means that most alternative multiple correction methods, such as [7] and [8] are discarded. Very few methods have confidence intervals, and even fewer are possible to integrate in the sample size calculation, due to their reliance on the p-values. One method that fulfills our conditions is Nyholt’s method for calculating the so-called effective number of independent tests in a set of tests with arbitrary covariance structure [14]. This method is frequently applied in high-dimensional genome testing, and has been refined by several, see e.g. [11, 5]. Nyholt’s method is simple, the effective number of independent tests is given by D.0.1 Monte Carlo simulation results with the Nyholt method Below the tables from Section 5 are repeated together with the result using Nyholts method as described above. As expected, the method improves the efficiency. This is shown by type I error rates and power closer to the intended rates under certain covariance structures. D.0.1 Monte Carlo simulation results with the Nyholt method ACKNOWLEDGMENTS The authors would like to thank Bob Wilson for feedback on earlier drafts of this paper. REFERENCES [1] BERGER, J. (2013). Statistical decision theory: foundations, concepts, and methods. Springer Science & Business Media. [2] DMITRIENKO, A., OFFEN, W. W. and WESTFALL, P. H. (2003). Gatekeeping strategies for clinical trials that do not require all primary effects to be significant. Statistics in Medicine 22 2387-2400. [3] DMITRIENKO, A., TAMHANE, A. C. and BRETZ, F. (2009). Multiple testing problems in pharmaceutical statistics. CRC press. [4] FABIJAN, A., GUPCHUP, J., GUPTA, S., OMHOVER, J., QIN, W., VERMEER, L. and DMITRIEV, P. (2019). Diagnosing Sample Ratio Mismatch in Online Controlled Experiments: A Taxonomy and Rules of Thumb for Practitioners. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. KDD ’19 2156–2164. Association for Computing Machinery, New York, NY, USA. https://doi.org/10.1145/3292500.3330722 [5] GALWEY, N. W. (2009). A new measure of the effective number of tests, a practical tool for comparing families of non-independent significance tests. Genetic Epidemiology: The Official Publication of the International Genetic Epidemiology Society 33 559–568. [6] GUILBAUD, O. (2014). Sharper Confidence Intervals for Hochberg-and Hommel-Related Multiple Tests Based On an Extended Simes Inequality. Statistics in Biopharmaceutical Research 6 123–136. [7] HOLM, S. (1979). A simple sequentially rejective multiple test procedure. Scandinavian journal of statistics 65–70. [8] HOMMEL, G. (1988). A stagewise rejective multiple test procedure based on a modified Bonferroni test. Biometrika 75 383–386. [9] KONG, L., KOHBERGER, R. C. and KOCH, G. G. (2004). Type I Error and Power in Noninferiority/Equivalence Trials with Correlated Multiple Endpoints: An Example from Vaccine Development Trials. Journal of Biopharmaceutical Statistics 14 893-907. PMID: 15587971. https://doi.org/10.1081/BIP-200035454 [10] LEHMACHER, W., WASSMER, G. and REITMEIR, P. (1991). Procedures for Two-Sample Comparisons with Multiple Endpoints Controlling the Experimentwise Error Rate. Biometrics 47 511–521. [11] LI, J. and JI, L. (2005). Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix. Heredity 95 221–227. [12] LIU, S. and LIU, A. (2022). The Lifecycle of Developing Overall Evaluation Criterion in AB Testing for Netflix Messaging. In Proceedings of the 26th International Conference on Evaluation and Assessment in Software Engineering 266–267. [13] NEUHÄUSER, M. (2006). How to deal with multiple endpoints in clinical trials. Fundamental & Clinical Pharmacology 20 515-523. [14] NYHOLT, D. R. (2004). A simple correction for multiple testing for single-nucleotide polymorphisms in linkage disequilibrium with each other. The American Journal of Human Genetics 74 765–769. [15] O’BRIEN, P. C. (1984). Procedures for Comparing Samples with Multiple Endpoints. Biometrics 40 1079–1087. [16] OFFEN, W., CHUANG-STEIN, C., DMITRIENKO, A., LITTMAN, G., MACA, J., MEYERSON, L., MUIRHEAD, R., STRYSZAK, P., BADDY, A., CHEN, K., COPLEY-MERRIMAN, K., DERE, W., GIVENS, S., HALL, D., HENRY, D., JACKSON, J. D., KRISHEN, A., LIU, T., RYDER, S., SANKOH, A. J., WANG, J. and YEH, C.-H. (2007). Multiple Co-primary Endpoints: Medical and Statistical Solutions: A Report from the Multiple Endpoints Expert Team of the Pharmaceutical Research and Manufacturers of America. Drug Information Journal 41 31–46. https://doi.org/10.1177/009286150704100105 Authors: (1) Mårten Schultzberg, Experimentation Platform team, Spotify, Stockholm, Sweden; (2) Sebastian Ankargren, Experimentation Platform team, Spotify, Stockholm, Sweden; (3) Mattias Frånberg, Experimentation Platform team, Spotify, Stockholm, Sweden. Authors: Authors: (1) Mårten Schultzberg, Experimentation Platform team, Spotify, Stockholm, Sweden; (2) Sebastian Ankargren, Experimentation Platform team, Spotify, Stockholm, Sweden; (3) Mattias Frånberg, Experimentation Platform team, Spotify, Stockholm, Sweden. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv available on arxiv

How Nyholt’s Method Makes Scientific Testing More Reliable

Authors:
(1) Albert Gu, Machine Learning Department, Carnegie Mellon University with Equal contribution (agu@cs.cmu.edu);
(2) Tri Dao, Department of Computer Science, Princeton University with Equal contribution (tri@tridao.me). Authors: Authors: (1) Albert Gu, Machine Learning Department, Carnegie Mellon University with Equal contribution (agu@cs.cmu.edu); (2) Tri Dao, Department of Computer Science, Princeton University with Equal contribution (tri@tridao.me). Table of Links Abstract and 1. Introduction Abstract and 1. Introduction 2 State Space Models 2 State Space Models 3 Selective State Space Models and 3.1 Motivation: Selection as a Means of Compression 3 Selective State Space Models and 3.1 Motivation: Selection as a Means of Compression 3.2 Improving SSMs with Selection 3.2 Improving SSMs with Selection 3.3 Efficient Implementation of Selective SSMs 3.3 Efficient Implementation of Selective SSMs 3.4 A Simplifed SSM Architecture 3.4 A Simplifed SSM Architecture 3.5 Properties of Selection Mechanisms 3.5 Properties of Selection Mechanisms 3.6 Additional Model Details 3.6 Additional Model Details 4 Empirical Evaluation and 4.1 Synthetic Tasks 4 Empirical Evaluation and 4.1 Synthetic Tasks 4.2 Language Modeling 4.2 Language Modeling 4.3 DNA Modeling 4.3 DNA Modeling 4.4 Audio Modeling and Generation 4.4 Audio Modeling and Generation 4.5 Speed and Memory Benchmarks 4.5 Speed and Memory Benchmarks 4.6 Model Ablations 4.6 Model Ablations 5 Discussion 5 Discussion 6 Conclusion, Acknowledgments and References 6 Conclusion, Acknowledgments and References A Discussion: Selection Mechanism A Discussion: Selection Mechanism B Related Work and B.1 S4 Variants and Derivatives B Related Work and B.1 S4 Variants and Derivatives B.2 SSM Architectures B.2 SSM Architectures B.3 Relationship to RNNs B.3 Relationship to RNNs B.4 Linear Attention and B.5 Long Context Models B.4 Linear Attention and B.5 Long Context Models C Mechanics of Selective SSMs C Mechanics of Selective SSMs D Hardware-aware Algorithm For Selective SSMs D Hardware-aware Algorithm For Selective SSMs E Experimental Details and Additional Results and E.1 Synthetic Tasks E Experimental Details and Additional Results and E.1 Synthetic Tasks E.2 Language Modeling E.2 Language Modeling E.3 DNA Modeling E.3 DNA Modeling E.4 Audio Details E.4 Audio Details E.5 Efficiency Benchmark E.5 Efficiency Benchmark 4.3 DNA Modeling Motivated by the success of large language models, there has been recent exploration into using the foundation model paradigm for genomics. DNA has been likened to language in that it consists of sequences of discrete tokens with a finite vocab. It is also known for requiring long-range dependencies to model (Avsec et al. 2021). We investigate Mamba as a FM backbone for pretraining and fine-tuning in the same setting as recent works on long-sequence models for DNA (Nguyen, Poli, et al. 2023). In particular, we focus on two explorations of scaling laws across model size and sequence length (Figure 5), and a difficult downstream synthetic classification task requiring long context (Figure 6). For pretraining, we largely follow a standard causal language modeling (next token prediction) setup for the training and model details (see also Appendix E.2). For the dataset, we largely follow the setup of HyenaDNA (Nguyen, Poli, et al. 2023), which uses the HG38 dataset for pretraining consisting of a single human genome with about 4.5 billion tokens (DNA base pairs) in the training split. 4.3.1 Scaling: Model Size In this experiment, we investigate the scaling properties of genomics foundation models with various model backbones (Figure 5 Left). Training. To advantage the baselines, we train on a short sequence length of 1024; as shown in Section 4.3.2, we expect results to favor Mamba even more at longer sequence lengths. We fix a global batch size of 1024, for a 12 4.3.2 Scaling: Context Length Results. Figure 5 (Right) shows that Mamba is able to make use of longer context even up to extremely long sequences of length 1M, and its pretraining perplexity improves as the context increases. On the other hand, the HyenaDNA model gets worse with sequence length. This is intuitive from the discussion in Section 3.5 on properties of the selection mechanism. In particular, LTI models cannot selectively ignore information; from a convolutional perspective, a very long convolution kernel is aggregating all information across a long sequence which may be very noisy. Note that while HyenaDNA claims to improve with longer context, their results do not control for computation time. 4.3.3 Synthetic Species Classification We evaluate models on a downstream task of classifying between 5 different species by randomly sampling a contiguous segment of their DNA. This task is adapted from HyenaDNA, which used the species {human, lemur, mouse, pig, hippo}. We modify the task to be significantly more challenging by classifying between the five great apes species {human, chimpanzee, gorilla, orangutan, bonobo}, which are known to share 99% of their DNA 4.4 Audio Modeling and Generation For the audio waveform modality, we compare primarily to the SaShiMi architecture and training protocols (Goel et al. 2022). This model comprises We consider replacing the S4+MLP blocks with Mamba blocks. Experiment details are in Appendix E.4. 4.4.1 Long-Context Autoregressive Pretraining 4.4.2 Autoregressive Speech Generation SC09 is a benchmark speech generation dataset (Donahue, McAuley, and Puckette 2019; Warden 2018), consisting of 1-second clips sampled at 16000 Hz of the digits “zero” through “nine” with highly variable characteristics. We largely follow the autoregressive training setup and generation protocol of Goel et al. (2022). Table 4 shows automated metrics of the Mamba-UNet model compared to a variety of baselines from Goel et al. (2022): WaveNet (Oord et al. 2016), SampleRNN (Mehri et al. 2017), WaveGAN (Donahue, McAuley, and Puckette 2019), DiffWave (Z. Kong et al. 2021), and SaShiMi. A small Mamba model outperforms the state-of-the-art (and much larger) GAN- and diffusion- based models. A larger model parameter-matched to the baselines further improves on fidelity metrics dramatically. 4.5 Speed and Memory Benchmarks 4.6 Model Ablations 4.6.1 Architecture Table 6 investigates the effects of the architecture (block) and its inner SSM layer (Figure 3). We find that • Among previous non-selective (LTI) SSMs, which are equivalent to global convolutions, performance is very similar. • Replacing the complex-valued S4 variant from previous work with a real-valued one does not affect performance much, suggesting that (at least for LM) real-valued SSMs may be a better choice when accounting for hardware efficiency. • Replacing any of these with a selective SSM (S6) significantly improves performance, validating the motivation of Section 3. • The Mamba architecture performs similarly to the H3 architecture (and seems slightly better when using a selective layer). We also investigate interleaving the Mamba block with other blocks such as MLP (a traditional architecture) MHA (a hybrid attention architecture) in Appendix E.2.2 4.6.2 Selective SSM Table 7 ablates the selective SSM layer by considering different combinations of selective ∆, B, and C parameters (Algorithm 2), showing that ∆ is the most important parameter due to its connection to RNN gating (Theorem 1). Table 8 considers different initializations of the SSM, which have been shown to make a large difference in some data modalities and settings (Gu, Goel, and Ré 2022; Gu, Gupta, et al. 2022). On language modeling, we find that simpler real-valued diagonal initializations (S4D-Real, row 3) instead of more standard complex-valued parameterizations (S4D-Lin, row 1) perform better. Random initializations also work well, consistent with findings from prior work (Mehta et al. 2023). Table 9 and Table 10 consider varying the dimension of the ∆ and (B, C) projections respectively. Changing them from static to selective provides the most benefit, while increasing the dimensions further generally improves performance modestly with a small increase in parameter count. Of particular note is the dramatic improvement of the selective SSM when the state size 푁 is increased, with over a 1.0 perplexity improvement for a cost of only 1% additional parameters. This validates our core motivation in Sections 3.1 and 3.3. Table 6: (Ablations: Architecture and SSM layer.) The Mamba block performs similarly to H3 while being simpler. In the inner layer, there is little dierence among dierent parameterizations of LTI models, while selective SSMs (S6) provide a large improvement. More specically, the S4 (real) variant is S4D-Real and the S4 (complex) variant is S4D-Lin. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Mamba’s Performance in DNA, Audio, and Speed Benchmarks

There is a recent perspective in The Transmitter, Why hasn’t genetics taught us more about schizophrenia?, stating that, "What should we conclude? I certainly don’t question the value of searching for genes that cause or confer susceptibility to disease and provide therapeutic targets. There have been far too many triumphs to leave any doubt that genome sequencing of various sorts will continue to be a centerpiece of disease-relevant research. On the other hand, in the realm of the brain, genetic approaches may be more valuable in studies of diseases in which cells appear to malfunction or die—generally neurodegenerative or neurological disorders—than those in which circuits appear to malfunction without obvious cellular pathology—generally behavioral or psychiatric conditions. Unraveling schizophrenia, bipolar disorder, depression, autism and so on will likely need to rely more heavily on other methods." "They might include human neural organoids to seek cellular and molecular changes, and neuroimaging to seek systems-level perturbations, as well as better biomarkers and identification of specific affected cell types. In the long run, results from these emerging methods could be combined with genetic data to teach us how schizophrenia happens and how to treat it effectively." What does it mean that "circuits appear to malfunction without obvious cellular pathology"? Circuits are neural circuits. So, if neural pathology is not the problem, what are the other options in the circuits? Electrical and chemical signals. There is no neural circuit without electrical and chemical signals. There is no other method in brain science that can be potent in addressing behavioral or psychiatric conditions without exploring electrical and chemical signals first, as the next best option because to rule them out would mean to have sought them out. How can the signals be approached? One option is to look at neurons in clusters. Neurons in clusters mean that electrical and chemical signals are in sets or in loops. There could be a set per cluster or several sets. The work, conceptually, of these sets is to configure specific information. This means that information is organized by sets of signals. Functions are preliminary information, obtained in the interactions of electrical and chemical signals. Functions include memory, feeling, emotion, and regulation of internal senses. These functions have several subdivisions. Attributes that qualify these functions include attention, awareness, subjectivity, and intent. Attributes are the states of respective signals, in sets, at the time of the interaction. So, towards solving psychiatric problems, the interactions and attributes of signals could be used to explain or construct a parallel to every condition in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR). For example, what is depression, by the architecture of electrical and chemical signals? It could be described by an attribute of chemical signals called principal spot, where a set of signals takes a higher than prioritized position, drawing electrical signals from other sets, reducing their measures of pre-prioritization. Simply, depression can be chemical signals at a principal measure [or spot], which then attracts summaries of electrical signals from several other sets, which may not have had anything to do with the set at the principal spot. This draw then reduces the optimal operation of those sets, resulting in experiences like heaviness, lack of motivation, sadness, and so forth. Other attributes that may be responsible include sequences, splits, and so forth. Every condition in the DSM-5-TR can be explained this way, in proximity to what might be ongoing within, by the operational components of the mind—the electrical and chemical signals. With this, it is possible to prospect therapies to remove the set from the principal measure by certain adjustments in an interval. This could also be similar to several other conditions. Even in seeking biomarkers, it is possible to use the interactions and attributes of electrical and chemical signals to place observable possibilities on measures that may represent biomarkers for some conditions. Genes are not the human mind. Neurons too are not the human mind. The goal is progress. What is ahead of neurons in how the brain works are the electrical and chemical signals. Exploring them could provide answers where genetics has struggled in aspects of brain science. There is a new paper in JAMA Psychiatry, Predicting Diagnostic Progression to Schizophrenia or Bipolar Disorder via Machine Learning, stating that, "Can diagnostic progression to schizophrenia or bipolar disorder be accurately predicted from routine clinical data extracted from electronic health records? In this cohort study of 24 449 patients aged 15 to 60 years with a total of 398 922 outpatient contacts to the Psychiatric Services of the Central Denmark Region between 2013 and 2016, progression to schizophrenia was predicted with higher accuracy than bipolar disorder, which proved to be a more difficult target. These findings suggest that detecting progression to schizophrenia through machine learning based on routine clinical data is feasible, which may reduce diagnostic delay and duration of untreated illness. These findings suggest that it is possible to predict diagnostic transition to schizophrenia and bipolar disorder from routine clinical data extracted from EHRs, with schizophrenia being notably easier to predict than bipolar disorder." There is a recent perspective in The Transmitter , Why hasn’t genetics taught us more about schizophrenia? , stating that, "What should we conclude? I certainly don’t question the value of searching for genes that cause or confer susceptibility to disease and provide therapeutic targets. There have been far too many triumphs to leave any doubt that genome sequencing of various sorts will continue to be a centerpiece of disease-relevant research. On the other hand, in the realm of the brain, genetic approaches may be more valuable in studies of diseases in which cells appear to malfunction or die—generally neurodegenerative or neurological disorders—than those in which circuits appear to malfunction without obvious cellular pathology—generally behavioral or psychiatric conditions. Unraveling schizophrenia, bipolar disorder, depression, autism and so on will likely need to rely more heavily on other methods." The Transmitter Why hasn’t genetics taught us more about schizophrenia? "They might include human neural organoids to seek cellular and molecular changes, and neuroimaging to seek systems-level perturbations, as well as better biomarkers and identification of specific affected cell types. In the long run, results from these emerging methods could be combined with genetic data to teach us how schizophrenia happens and how to treat it effectively." What does it mean that "circuits appear to malfunction without obvious cellular pathology"? Circuits are neural circuits. So, if neural pathology is not the problem, what are the other options in the circuits? Electrical and chemical signals. There is no neural circuit without electrical and chemical signals. There is no other method in brain science that can be potent in addressing behavioral or psychiatric conditions without exploring electrical and chemical signals first , as the next best option because to rule them out would mean to have sought them out. exploring electrical and chemical signals first How can the signals be approached? One option is to look at neurons in clusters. Neurons in clusters mean that electrical and chemical signals are in sets or in loops. There could be a set per cluster or several sets. The work, conceptually, of these sets is to configure specific information. This means that information is organized by sets of signals. Functions are preliminary information, obtained in the interactions of electrical and chemical signals. Functions include memory, feeling, emotion, and regulation of internal senses. These functions have several subdivisions. Attributes that qualify these functions include attention, awareness, subjectivity, and intent. Attributes are the states of respective signals, in sets, at the time of the interaction. So, towards solving psychiatric problems, the interactions and attributes of signals could be used to explain or construct a parallel to every condition in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR). For example, what is depression, by the architecture of electrical and chemical signals? It could be described by an attribute of chemical signals called principal spot, where a set of signals takes a higher than prioritized position, drawing electrical signals from other sets, reducing their measures of pre-prioritization. Simply, depression can be chemical signals at a principal measure [or spot], which then attracts summaries of electrical signals from several other sets, which may not have had anything to do with the set at the principal spot. This draw then reduces the optimal operation of those sets, resulting in experiences like heaviness, lack of motivation, sadness, and so forth. Other attributes that may be responsible include sequences, splits, and so forth. Every condition in the DSM-5-TR can be explained this way , in proximity to what might be ongoing within, by the operational components of the mind—the electrical and chemical signals. With this, it is possible to prospect therapies to remove the set from the principal measure by certain adjustments in an interval. This could also be similar to several other conditions. Even in seeking biomarkers, it is possible to use the interactions and attributes of electrical and chemical signals to place observable possibilities on measures that may represent biomarkers for some conditions. Genes are not the human mind. Neurons too are not the human mind. Every condition in the DSM-5-TR can be explained this way The goal is progress. What is ahead of neurons in how the brain works are the electrical and chemical signals. Exploring them could provide answers where genetics has struggled in aspects of brain science. There is a new paper in JAMA Psychiatry, Predicting Diagnostic Progression to Schizophrenia or Bipolar Disorder via Machine Learning , stating that, "Can diagnostic progression to schizophrenia or bipolar disorder be accurately predicted from routine clinical data extracted from electronic health records? In this cohort study of 24 449 patients aged 15 to 60 years with a total of 398 922 outpatient contacts to the Psychiatric Services of the Central Denmark Region between 2013 and 2016, progression to schizophrenia was predicted with higher accuracy than bipolar disorder, which proved to be a more difficult target. These findings suggest that detecting progression to schizophrenia through machine learning based on routine clinical data is feasible, which may reduce diagnostic delay and duration of untreated illness. These findings suggest that it is possible to predict diagnostic transition to schizophrenia and bipolar disorder from routine clinical data extracted from EHRs, with schizophrenia being notably easier to predict than bipolar disorder." JAMA Psychiatry, Predicting Diagnostic Progression to Schizophrenia or Bipolar Disorder via Machine Learning

Stories, News & Mentions about GENOME

Categories

Trending Topics

Stories, News & Mentions about GENOME

Genome’s Stories on HackerNoon

Genome's News & Mentions

Categories

Trending Topics