In progress
A Bayesian claim-evidence framework for CYP inhibition
I am working through how different evidence sources relate to CYP assay endpoints, without treating every assay or model output as interchangeable ground truth.
Current focus
The framework is being developed around two endpoint families: direct-inhibition pIC50 for CYP1A2, CYP2C9, CYP2D6, and CYP3A4; and time-dependent inhibition derived from matched +NADPH and -NADPH assay arms for CYP3A4 and CYP2D6.
Current challenge data, rules, dates, and submission details are maintained on the OpenADMET CYP Challenge Hugging Face Space.