Renumbering a TCR-pHLA complex changes HLA residues to negative values

[victorlr1992]

I just noticed that when I submit a TCR-pHLA complex for PDB renumbering, the HLA residues are renumbered. HLA residues starts at (very high) negative values. Is there a workaround or a way of ignoring certain chains?

[ALGW71]

Hi @victorlr1992 - thanks for this. I am making a lot changes in the next few days and one of those is to change the cutoff in TCR mode, as the score threshold for the TCR models is different from the antibody models, this may solve your issue.

However a feature that allows the user to pick the PDB chains they want to be numbered will be much better, so will keep this issue open until we have that.

Longer term we are thinking about HLA numbering, but that will take a bit longer.

Will update you soon!

[ALGW71]

Hi @victorlr1992 sorry about the delay on this (crazy year), we are working on numbering MHC as well.

A simple hack for your problem was to change the cutoff scores see #106.

A more complicated solution is this: #100

If you want to number MHC as well, please try that branch and leave any comment on the PR. I think it is mostly working, the issue is the classifier is heavily corrupted by MHC (highly diverse sequences) and thus labels many things as MHC, and tries to number them... I am trying to figure out a solution to this that doesn't introduce many problems into the whole codebase...