Importing TCR Sequences Without V, J, CDR3 Names for Predicted PAE Binding Prediction

phbradley / TCRdock

python tools for TCR:peptide-MHC modeling and analysis

MIT License

66 stars 9 forks source link

Importing TCR Sequences Without V, J, CDR3 Names for Predicted PAE Binding Prediction #21

Open kanyuanhao opened 4 months ago

kanyuanhao commented 4 months ago

Hi, it's an awesome tool. I have some PDB files of TCR:pMHC complexes, therefore I have the sequences for each TCRα, TCRβ, MHC, and peptide. However, I don't have the names for V, J, CDR3, etc. Is it possible to import these sequences to get the predicted PAE for binding prediction?

Alternatively, how can I determine the names of V, J, and CDR3 from these sequences? I tried using IgBLAST but wasn't successful (Only got V but not J and CDR3).

Thank you very much!

phbradley commented 4 months ago

Hi there, great question! I am embarrassed to admit that I don't know how to do that with web tools. I thought IGblast would parse out the J gene and CDR3 -- maybe they are hiding somewhere in the output? The CDR3 you can generally find by eye by looking for the second cysteine and continuing up to and including the F/Y/W that comes right before a GXG motif.

kanyuanhao commented 3 months ago

@phbradley Thank you very much. I have no idea why IgBlast couldn't generate the results, but I will follow your advice. ❤️