The postdoctoral fellow will achieve the following learning goals: (1) develop rigorous and reproducible statistical and machine learning methods for integrating multi-modality cancer datasets, with strong benchmarking and uncertainty awareness, and deliver these methods as well documented computational tools; (2) build AI pathology models that convert tissue morphology into quantitative features to support downstream molecular interpretation, including deconvolution and harmonization approaches for robust comparison across patients, cohorts, and tissue types; (3) create agentic AI workflows that automate analysis from data ingestion and quality control to interpretation and report generation, with emphasis on transparency, auditability, and scalability on high performance computing systems; (4) conduct integrative modeling of 3D genome organization and cross platform cell surface protein measurements to improve gene regulation insight and cell type and state characterization; (5) develop professional skills through structured mentorship in manuscript writing, scientific communication, and career development applications, including K99 R00 and Damon Runyon. This computational postdoctoral fellow candidate is expected to leverage single-cell/bulk-cell multi-omics, spatial omics, and pathological imaging data to reveal the cancer-specific mechanisms underlying the differential efficacies and toxicities of treatments across patients.