@laurenlporter.bsky.social
Investigates shapeshifting proteins with computation and experiment. Tenure-track. Mother of two. Foodie. Yoga lover. Opinions are mine.
Sure. How would you suggest addressing it?
Doing my best to challenge the "because AI" argument! Science is all about explanations. If a neural network cannot be explained, seems to me it's not science. It's excellent engineering.
Doing my best to challenge the "because AI" argument! Science is all about explanations. If a neural network cannot be explained, seems to me it's not science. It's excellent engineering.
December 14, 2023 at 11:49 PM
Sure. How would you suggest addressing it?
Doing my best to challenge the "because AI" argument! Science is all about explanations. If a neural network cannot be explained, seems to me it's not science. It's excellent engineering.
Doing my best to challenge the "because AI" argument! Science is all about explanations. If a neural network cannot be explained, seems to me it's not science. It's excellent engineering.
Agreed. I have never heard John Jumper or anyone else on the AF2 team overstate its capabilities. The confusion stems from other sources: misunderstandings of structural biology and overconfidence in AI.
December 14, 2023 at 11:32 PM
Agreed. I have never heard John Jumper or anyone else on the AF2 team overstate its capabilities. The confusion stems from other sources: misunderstandings of structural biology and overconfidence in AI.
Thanks for the warm welcome 😀. It's good to be here.
December 14, 2023 at 5:46 PM
Thanks for the warm welcome 😀. It's good to be here.
Agreed-- my lab uses AF2 all of the time (with caution) to generate hypotheses. Correct predictions do not imply learned folding physics.
December 14, 2023 at 2:44 PM
Agreed-- my lab uses AF2 all of the time (with caution) to generate hypotheses. Correct predictions do not imply learned folding physics.
These results lead me to believe that:
1). While AF2 builds great models, it does not equal experimental structure determination.
2). AF2 has not learned much protein folding physics. Rather, it has likely learned what good protein structures look like.
1). While AF2 builds great models, it does not equal experimental structure determination.
2). AF2 has not learned much protein folding physics. Rather, it has likely learned what good protein structures look like.
December 14, 2023 at 11:53 AM
These results lead me to believe that:
1). While AF2 builds great models, it does not equal experimental structure determination.
2). AF2 has not learned much protein folding physics. Rather, it has likely learned what good protein structures look like.
1). While AF2 builds great models, it does not equal experimental structure determination.
2). AF2 has not learned much protein folding physics. Rather, it has likely learned what good protein structures look like.
Together, these results suggest that AF2 has more to learn about protein energy landscapes. They also suggest that AF2 sometimes struggles to associate amino acid sequences with their correct conformations, as it did for BCCIP-alpha (correct fold on right):
December 14, 2023 at 11:53 AM
Together, these results suggest that AF2 has more to learn about protein energy landscapes. They also suggest that AF2 sometimes struggles to associate amino acid sequences with their correct conformations, as it did for BCCIP-alpha (correct fold on right):
Further, AF2 and AF-cluster failed to predict new fold switchers discovered after AF2.3.1 was trained:
December 14, 2023 at 11:52 AM
Further, AF2 and AF-cluster failed to predict new fold switchers discovered after AF2.3.1 was trained:
AF2 based methods performed worse when predicting fold switchers outside of the training set. For instance, AF-cluster predictions could not distinguish between sequence-diverse fold-switching and single-folding RfaH homologs, and all helical predictions had low confidence.
December 14, 2023 at 11:52 AM
AF2 based methods performed worse when predicting fold switchers outside of the training set. For instance, AF-cluster predictions could not distinguish between sequence-diverse fold-switching and single-folding RfaH homologs, and all helical predictions had low confidence.
Consistently, AF2's structure module--hypothesized to have learned folding physics--could not discriminate between low and high energy conformations of fold switchers. (TM-scores on x- and y- axes).
December 14, 2023 at 11:51 AM
Consistently, AF2's structure module--hypothesized to have learned folding physics--could not discriminate between low and high energy conformations of fold switchers. (TM-scores on x- and y- axes).
AF2 confidence metrics selected against diverse experimentally observed conformations, particularly alternative folds (p < 8.1*10-4, one-sided binomial test), in favor of experimentally unobserved conformations
December 14, 2023 at 11:50 AM
AF2 confidence metrics selected against diverse experimentally observed conformations, particularly alternative folds (p < 8.1*10-4, one-sided binomial test), in favor of experimentally unobserved conformations
We tested 5 AF2-based methods on 93 fold-switching proteins likely in AF2’s training set, >280,000 predictions total. Each method predicted <20% of fold switchers and favored experimentally unobserved folds (other) over experimentally observed alternative folds (Fold2).
December 14, 2023 at 11:50 AM
We tested 5 AF2-based methods on 93 fold-switching proteins likely in AF2’s training set, >280,000 predictions total. Each method predicted <20% of fold switchers and favored experimentally unobserved folds (other) over experimentally observed alternative folds (Fold2).