Ugh, sorting through some old alpha-fold predictions and find something that looks quite strong but have no clue why I was looking at that protein pair. *sigh*

#PIlife #StructurePredictions

Because it is the start of the month, I'm indexing my notebook and seeing that there is a whole set of really exciting protein multimer predictions that still need to be done.

Of course, our clusters are down for summer maintenance this week. ugh.

I'm so tempted to use the browser-based colabfold, but I should just wait for the clusters.

I'm very impatient.

#AcademicChatter #LabLife #AlphaFold2 #StructurePredictions

@askennard

In this preprint, we identified a conserved or suspiciously structurally similar interface between a large highly conserved well-folded protein and its fast-evolving co-factors in yeast and humans. Some co-factors are large, but all of them consist primarily of disordered regions.

https://www.biorxiv.org/content/10.1101/2023.08.24.554671v2

We are now using alpha fold to predict protein-protein interactions amongst proteins that are linked to these ones via phenomics.

#StructurePredictions #Phenomics #alphafold

Note in point.

What does it mean when only your rank 1 model suggests an interaction between two proteins with alphafold multimer? Is this garbage?

#AlphaFoldMultimer
#AlphaFold
#AlphaFold2
#ProteinPrediction
#StructurePredictions
#ProteinPredictionHelp

When submitting a manuscript with alphafold2-based predictions, what all should I provide as supplemental data?

I'm thinking the rank 1 model PDB and json file, and then for all of the models (5 in my case) the pae.png and plddt.png.

Would someone want to see more than this or less than this?

#AlphaFold #AlphaFold2 #StructurePredictions #DeepMind