Atomwise (YC W15) | Infrastructure, DevOps, Machine Learning | San Francisco | Full-time | Onsite | https://www.atomwise.com/careers/
Atomwise Inc. patented the first deep learning technology for structure-based small molecule drug discovery. This AI technology harnesses millions of data points and thousands of protein structures to solve problems that a human chemist would take many lifetimes to solve. Atomwise has partnered with some of the world’s largest pharmaceutical and agrochemical companies, and with more than 50 leading academic institutions and hospitals, to tackle the challenges of discovering and developing better drugs and chemicals. Recently, Atomwise raised $45 million from leading venture capital firms to support the development and application of its AI technology.
Atomwise (YC W15) | Infrastructure, DevOps, Machine Learning | San Francisco | Full-time | Onsite | https://www.atomwise.com/careers/
Atomwise Inc. patented the first deep learning technology for structure-based small molecule drug discovery. This AI technology harnesses millions of data points and thousands of protein structures to solve problems that a human chemist would take many lifetimes to solve. Atomwise has partnered with some of the world’s largest pharmaceutical and agrochemical companies, and with more than 50 leading academic institutions and hospitals, to tackle the challenges of discovering and developing better drugs and chemicals. Recently, Atomwise raised $45 million from leading venture capital firms to support the development and application of its AI technology.
That's a very useful tool! Part of the reason that we launched the Atomwise AIMS program http://www.atomwise.com/aims-awards/ was to address the cost of compounds, in addition to the cost of equipment. We use our deep neural networks to pick 72 compounds (out of millions) that we buy, QC, plate, and ship to the PI for free.
Atomwise (YC W15) | San Francisco | Full time, ONSITE | Deep Learning, Computational Chemistry
Atomwise uses deep neural networks to help discover new medicines. Our customers are top researchers at institutions such as Merck and the Dana Farber Cancer Institute (see http://www.atomwise.com/clients/). We're backed by science-heavy VCs, including Data Collective, Khosla Ventures, and DFJ. Our work tackles some of the biggest problems of our time: cancer, multiple sclerosis, malaria, ebola, and antibiotics for drug-resistant bugs. We’ve already shown that modern machine learning can set a new bar for predictive accuracy in structure-based drug design, and we want your help in pushing that accuracy even further.
We’re looking both for people with machine learning expertise, and for people with computational biology/chemistry expertise. If you’ve got both, all the better! Please see our full job descriptions here: http://www.atomwise.com/careers/
Atomwise (YC W15) | San Francisco | Full time, ONSITE | Deep Learning, Computational Chemistry
Atomwise uses deep neural networks to help discover new medicines. Our customers are top researchers at institutions such as Merck and the Dana Farber Cancer Institute (see http://www.atomwise.com/clients/). We're backed by science-heavy VCs, including Data Collective, Khosla Ventures, and DFJ. Our work tackles some of the biggest problems of our time: cancer, multiple sclerosis, malaria, ebola, and antibiotics for drug-resistant bugs. We’ve already shown that modern machine learning can set a new bar for predictive accuracy in structure-based drug design, and we want your help in pushing that accuracy even further.
We’re looking both for people with machine learning expertise, and for people with computational biology/chemistry expertise. If you’ve got both, all the better! Please see our full job descriptions here: http://www.atomwise.com/careers/
Thank you! Personally, I find it very exciting to be working on these problems.
With respect to boosting, we have more investigation to do, of course; the tricky issue with the biological domain is that we know the underlying data is incredibly noisy. How to walk the line of extracting maximum predictive performance without overfitting is the challenge, since we know that a lot of the raw data points are unreliable. Any algorithm we use has to be able to handle this scenario deftly.
Today, those tests are done physically. But, you're right: if you have a good system to tell if a molecule will stick to a given protein, there's no reason to constrain your tests to the protein you want to hit. You can also predict whether the molecule will go around sticking to necessary proteins in the heart (e.g., hERG channel), liver (e.g., cytochrome P450), kidney, brain, etc. Internally, we have a panel of a couple of hundred proteins against which we can predict these kinds of off-target toxicities.
The typical timeline to get an actual cure all the way through the drug discovery pipeline is about 14 years. While we haven't been around long enough for that, we have had our algorithmic predictions validated by follow up physical experiment. This was even for very different diseases, e.g. multiple sclerosis and drug-resistant antibiotics.
Also, as I described in my above answer to et2o, we do large retrospective tests to evaluate our predictive accuracy.
The typical input to the neural network is the 3D structure of the molecule and of the protein. The model works by detecting patterns in the pair of protein and the drug that correlate with binding, e.g. hydrogen bonding, halogen bonding, cation-pi, pi-pi interactions, etc. But these are complicated to encode manually, given all of the factors that affect binding strength: distance, angle, water mediated effects, resonance, (de)stabilizing environmental charges, etc. That's why we need the neural net: you can think of it as the network automatically deriving the best pharmacophoric features to maximally explain which training examples bind and which ones don't, and then the prediction step is looking for the presence or absence of those patterns in new protein-ligand pairs.
We evaluate our models both retrospectively and prospectively. For example, the DUD-E benchmark (http://dude.docking.org/) gives us an assessment of our performance over more than a million individual predictions, comprising many diseases and many biological classes (GPCR, nuclear receptor, enzyme, etc). It begins with 102 disease proteins and, for each one, has a set of molecules that bind to the protein and a set that don't. We shuffle those sets together and ask the neural net to "pick the aces out of the deck". Separately, we perform prospective evaluations, for settings where no one knows the right answer, and run the experiment to confirm the predictions.
I agree with you that the proper selection of targets is critical, as is the mapping between drug target and disease. For us, however, this is easy: we work with smart biologists! If you have any, please send them our way!
Finally, I agree with your point that biology is not designed to be understood by people. That said, molecular binding is fundamental enough that we could think of it as an example of physics rather than biology. And theory works so well for physics that, in many a physics lab, if an experiment disagrees with theory then the first step is to double-check the experiment for errors. The trick is to scale that up to larger systems. Semi-relevant: http://www.smbc-comics.com/?id=2272
New medical discoveries aside, we're seeing self-driving cars and speech recognition that runs on a cell phone. I grew up reading about those kinds of things in Asimov, so I personally find the progress pretty exciting.
Over the past few years, there's been a huge increase in the amount of data available for this kind of machine learning. We curate our data from a number of private and public sources. For example, as part of my doctoral work (http://en.wikipedia.org/wiki/SCRIPDB), I learned how to parse chemical information out of U.S. Patent data, which is public domain. That said, if you're interested in working on something like this and need a quick million data points, I'd point you to PubChem as a first step: https://pubchem.ncbi.nlm.nih.gov/
As you might expect, there are trade-offs, and it's a question of picking the right tool for the job.
My understanding of D.E. Shaw's approach is that they're doing molecular dynamics, i.e. simulation. You get to watch the motion of every atom in the system. That allows for a close investigation of a given protein's movements, which is great especially if you're trying to learn about its biology. Unfortunately, it is rather computationally expensive; while I don't know DESRES's latest stats, I've seen reports on large parallel MD simulations completing about once per day.
In contrast, we've posed the question of binding as a machine learning problem. Neural networks are computationally expensive to train, but make predictions quickly. Our system can assess millions of protein-drug pairs per day, since we're not simulating the motion of every atom. You don't get to watch what each atom is doing, but you get insight into the behavior of lots of potential medicines.
I’m the cofounder and CEO of Atomwise and, since this is Hacker News, I thought I’d cover the technical details a bit more: We run deep neural networks on one of the biggest supercomputers (#74 in the world, http://www.top500.org/site/50424) to predict whether a molecule will stick to a disease target (its “binding affinity”). Understanding the binding affinity of a molecules is one of the essential questions in finding new medicines; it comes up over and over in the drug discovery pipeline, including in hit discovery, toxicity prediction, and personalized medicine.
Our goal is to bring to medicine discovery the same kind of incredible efficiency gains that computation gave us in aerospace and mechanical engineering design. Today, people have to physically synthesize and physically test molecules to figure out how they’re going to behave. That’s incredibly laborious, expensive, and time consuming. We’re able to get the same results, but in days instead of months or years.
Given all of the new and re-emerging diseases we’re encountering (such as Ebola, measles, malaria, and drug-resistant infections, to name a few that we’ve worked on), I think our species needs all of the help we can get in finding new medicines. I’m happy to answer questions about what we’re doing, or the challenges we encounter when we take deep learning algorithms out beyond image classification.
Atomwise Inc. patented the first deep learning technology for structure-based small molecule drug discovery. This AI technology harnesses millions of data points and thousands of protein structures to solve problems that a human chemist would take many lifetimes to solve. Atomwise has partnered with some of the world’s largest pharmaceutical and agrochemical companies, and with more than 50 leading academic institutions and hospitals, to tackle the challenges of discovering and developing better drugs and chemicals. Recently, Atomwise raised $45 million from leading venture capital firms to support the development and application of its AI technology.