3 ms·
Trading accuracy for speed using language based models: > Meta’s network, called ESMFold, isn’t quite as accurate as AlphaFold, Rives’ team reported earlier th
by tromp 4y ago
Trading accuracy for speed using language based models:
> Meta’s network, called ESMFold, isn’t quite as accurate as AlphaFold, Rives’ team reported earlier this summer2, but it is about 60 times faster at predicting structures, he says. “What this means is that we can scale structure prediction to much larger databases.”
> Burkhard Rost, a computational biologist at the Technical University of Munich in Germany, is impressed with the combination of speed and accuracy of Meta’s model. But he questions whether it really offers an advantage over AlphaFold’s precision, when it comes to predicting proteins from metagenomic databases. Language model-based prediction methods — including one developed by his team3 — are better suited to quickly determine how mutations alter protein structure, which is not possible with AlphaFold. “We will see structure prediction become leaner, simpler cheaper and that will open the door for new things,” he says.
- dekhn 4y agoStructure prediction is embarassingly parallel, and rarely requires any specific protein to be predicted- which can be done at any time, including precomputed- in an extremely short period of time. DeepMind has no trouble getting the inference time at Google to compute predictions using their model for any protein of interest or the largest database. Thus, a faster but more inaccurate system is not really desirable unless it really does provide another feature, such as (as Rost says) predictions on mutations alter structure. But if alphafold works, then it would indeed be capable of predicting how mutations alter structure.