Changes between Version 18 and Version 19 of NIAID2024


Ignore:
Timestamp:
Aug 1, 2024, 1:37:12 PM (15 months ago)
Author:
Elaine Meng
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • NIAID2024

    v18 v19  
    8585== 8/1/24 ==
    8686
     87Elaine, Eric, Zach, Piet, Scooter, Tom G; Phil, Darrell, Mike Bopf, Kristen
     88
     89Re NIH3D pipeline:
     90
     91Phil: what was the high-confidence cutoff for !AlphaFold?  Eric: 50, could easily be changed. Phil: could the printable output only show the single biggest piece? Eric: doable.  Phil: another separate rule, is that if high confidence is <17 residues total, do not create output (visible or printable). Elaine: for the biggest piece, what about separate domains that are are not contiguous covalently but that are in contact?  Scooter: let's start with just the covalent connectivity because it's simpler.  Eric: could do some kind of distance-based contact test, NxN where N is the number of contiguous pieces. Phil: can we use the "PAE domains"? TomG: probably too complicated. Could just do NxN contacts where N is the number of residues and then take biggest connected/contacting piece.  Elaine: get a contact RIN and then use the largest connected component.
     92
     93Kristen: Eric's previous fix for the virus chain-names issue looks good.
     94
     95Phil: question about contacts command, can we list only residue-residue contacts?  Elaine: no, is atom-atom; you can use contacts to select atoms and then list selected residues, but it would lose the residue pairing information. Eric: can add an option to the command.  Phil: sounds good.
     96
     97Phil: plans to incorporate AlphaFold3?  TomG: their website only allows limited calculations and there is no API for programmatic access. However, can upload a json to set up a run on their server.  So one potential thing ChimeraX could do is generate that json file, but it may be difficult especially for ligands, and not worth it with the current restrictions on what calculations can be done.  So I'm inclined to wait and hope that a less restrictive version becomes available.  Phil: I miss the interactive PAE.  TomG: currently (v1.8+) you can download the results and then open the AlphaFold3 PAE data and structure in ChimeraX. Phil: OK, then I'm satisfied!
     98
    8799== 7/18/24 ==
    88100