orintorynchus

Not exactly "finding out", but rather estimating. It's all about prioritizing testing through detecting suspiciously "good" and unexpected outliers popping up.

That's the key bit here. Supervised learning is not applicable here and any sort of fitting to known labels is doomed to fail. The system is not calibrated, tuned or parameterized. The ML part learns in a self-supervised manner what are the spike sequence features important for a successful (proliferating) coronavirus (or - more correctly - learns low-dimensional embeddings of multi-point interactions/co-occurrences of different amino acids in spike proteins). The rest is based on either frequentist statistics, or computational biochemistry. At no point in training EWS any information about certain sequences belonging to High Risk Variant classes (Variant of Concern, Variant of Interest etc.) is fed to EWS.

The idea is to introduce a bit of sanity into the current world, to calm down overreactions to every new, scary looking variant. And on the flip side - to identify the truly scary ones early on, so that proper measures can be taken.

But the idea above directly follows from the work in the paper.

We are actively looking for motivated colleagues. Feel free to look at the page above or drop us an email at hello[at]instadeep.com. We are also happy to make new friends and work together on exciting projects - same email as above works.

The method, in its current incarnation, is detecting dangerous variants and not foreseeing them. We can use the same approach to forecast plausible developments, but there are so many latent variables (intrapatient evolution, mobility, vaccination status, restriction compliance), that it is more of an informed "What If?" exercise, than a true prediction. Same as with many good stories on predicting the future, we can only see what is likely to happen, not what will necessarily happen... Out of many paths only a few will be explored in the end :-)

I am sure we will, in due time. If you have any particular hopes or wishes, we will try to accommodate as much as we can.

This is something that amazes me. Any time a true High Risk Variant appears, it is clear as day in the system. This was the case with Lambda, Mu (which we predicted to have a limited propensity to proliferate), and now with Omicron. However, there are lineages, which are prospectively dangerous, that we detect. As the classification is relative, there is no fixed threshold beyond which we would call for alert. In the evaluation, we have been looking at 20 sequences per week, as this was roughly the testing capacity of our partners. Going for ONE sequence a week makes us detect some variants a bit later, but still. The sequences and lineages we predicted to be of interest were predominantly spreading afterwards. Some were just a blip on the radar, though. We aimed at sensitivity and not missing ominous signs, rather than specificity. NB: Each week there are thousands (now 12k+) new sequence variants. Out of them we consider 20. And detect most of the variants as early on as on the first day.

The system is constantly learning, so it "retunes" itself. We can infer epitopes from structure, but we found that data derived from known complexes is sufficient for our purpose.

Current version of EWS does not explicitly account for glycosylation. It is implicitly handled by the ML models, though.

For specificity, it is difficult to estimate. We know that for each of the named Variants of Concern, the signal from EWS was unmissable. Considering, that any new vaccine would need to go through a stringent approval process, I don't think that EWS should be a major determining factor in the process. However, it can certainly help resolve the doubts.

What we hope for is rather informing the public policy. Now, any time a scary looking variant is sequenced, there is some public commotion, uncertainty about the future repercussions. We want to be able to gauge the appropriate level of concern in these situations. It is not an all-knowing oracle, but rather a way to distill the insights from prior observations and simulations, the same way a human expert would do. This being said, we believe that the insights provided by EWS can be of use in designing new vaccines, as well as deploying the existing ones most effectively.

Well, there is a lot of overlapping nomenclature. Here by variant we understand any sample, which is not identical sequence-wise to sequences seen before. Most of the observed mutations are innocuous and do not lead to a potential new lineage (what better fits the definition of variant above). However, it is difficult to tell by eye - or even with the standard computational tools - if new mutation is (a) deleterious (harms the virus' capacity to spread), (b) neutral, or (c) fitness enhancing.

It is not random. InstaDeep has designed, developed, built, benchmarked the method. BioNTech performed experimental validation, provided expert insight into the problem domain. Hence, authors are segmented by contribution areas.

I am the second author of the paper. We have been putting our head on the line for the last half a year. We detected Lambda, Mu (with a caveat, that we did not consider it competitive) and Omicron - all blindly.

We have been verifying all our predictions experimentally, post factum. And the method is purely data driven - with no fitting to the experiments or observations.

The first author came up with the approach, participated in analysis and got his hands dirty as everyone else. While he is a CEO of a 160+ person company, this has been a labor of love for all of us, done to a large extent in the evenings, during weekends and holidays. It is indeed an unusual situation. But this was not a regular project and InstaDeep is not a regular company either.