Will we be able to generate all stable protein conformations from a sequence by 2027?
8
34
260
Dec 31
24%
chance

Right now, Alphafold2 generates only the 3D structure for one of many possible protein conformations, with no way of easily guessing the alternative stable confirmations or the free energy associated to them.

As important as structure prediction is, do you think it will be upgraded for the prediction of free energy landscapes before the beginning of 2027?

There must be a scientific consensus, as there is with the effectiveness of alphafold2. Possibly through comparison with the results of molecular dynamics simulations and/or experiments.

Get Ṁ200 play money
Sort by:

I think it is quite possible, perhaps even likely to predict free energy landscapes. Which is very much not the same as "generate all possible protein conformations", an impossible task!

@Zozo001CoN I have changed the question from "possible" to "stable", which may be less ambiguous. Would not predicting the free energy landscape imply that it is possible to know all stable conformations?

@EnriqueAlanis well no - the landscape is a collection of relatively low energy conformations (and pathways among them), that excludes theoretically stable local minima which are unreachable from the ground state

More related questions