For T cells you could scan peptides spanning the amino acid changes for different HLA types. There are mature computational tools for that. Testing for B cell eptiopes will be more difficult IMO.

as far as I know there are mostly old school tools for this:
tools.iedb.org/main/bcell/

You may already be familier I think. For example, these tools try to find hydrophpic regions.

I'm curious how this might be helpful to you? Do you mind sharing the goal?

what level of information do you have? BCR only? or BCR and target/antigen? or target/antigen only?

100% agree; would add a corollary to (2) to include immune infiltrate.

+ (4) Incomplete understanding of the necessary T cell repertoire. At the end of the day, if there is no T cell that can recognize neoantigens and get activated, I'm hard pressed to imagine how we get the immune response started

TMB = measures a tumor's antigen potential -- I think we'll get better as we improve measures of clonal+neoantigens.

PD-L1 = IMO is actually a poor measure of tumor extrinsic properties ie T cell -> Ifng -> tumor PD-L1. I think we would be better off directly measuring T cell cytotoxic activity.

Late to the party (+I am biased), but I think combining immune features and tumor features is likely to be the best path toward creating an integrative, predictive biomarker for checkpoint response.

Also for the cancer TCR data sets -- Have you found a sustainable solution or near-sustainable solution to expand datasets? I think the current spanshot is useful as-is but curious if you have something in place to help expand as new data gets generated/published.

This looks really cool and seems like a great application of jointly embedding TCR+GEX. Also congrats on a great software setup. so I had a question about some of the data you've organized. I saw you had a repo of TCR-collected libraries; have you also collected data from BCRs?

lol if we're willing to go down that road - might as well co-opt the 🖖 emoji as an antibody symbol :) the thumb is the Fc receptor.

lol - maybe a parent enlisted their kid to help write copy for the mail adverts? the penmanship is great but agreed - you can tell the curves are too joyful and the lines too hopeful. It's got that youthful vibe that says “I haven’t had to deal with a difficult customer yet"

That emoji always reminds me of the "cancer immunogram" 😀
www.science.org/doi/10.1126/...

I was searching on here to see what the immunology conversation was like—so glad to see this! btw agreed! The whiplash from reading "impaired antigen presentation is the basis for ICB resistance" to "well, actually some patients do respond to ICB despite HLA loss" happened in a crazy short time!

Thanks for giving one of those to me!