This is a fantastic challenge. Thank you for organizing.
Our work modeling survival in NSCLC patients treated with ICI has lead us to nominate several clinically actionable annotations that are currently unavailable.
https://pubmed.ncbi.nlm.nih.gov/33398162/
In our data sets, the effect size of these clinical variables (liver mets/ NLR, etc) is far larger than gene signatures, TMB, PD-L1, TCR clonality.
Even if its outside the scope of the current challenge, can we have a conversation about adding some overlooked variables that may improve prediction?
Best,
Mike
Created by Michael Green mdgreen004 Dear @mdgreen004 ,
Thanks for your interest in the Anti-PD1 Response Prediction DREAM Challenge. Depending on the results of this Challenge and available clinical features we may reach out regarding directions of post-challenge analysis.
Kind regards,
Mike