Steven Pastor

WORK HISTORY

EDUCATION

SKILLS

ABOUT STEVEN PASTOR

BS Biology, MS Biology (Molecular), PhD Biomedical EngineeringCore Competencies: Structural Biology & Proteomics: AlphaFold • PyMOL • ChimeraX • MSFragger (via FragPipe) • DIA-NN • Spectronaut • PTM discovery • HLA-binding prediction • variant-to-structure pipelines Multi-omics & Sequencing: WGS/WES (Illumina) • Long-read (ONT/PacBio) • Optical mapping • scRNA-seq • scATAC-seq • CITE/TCR-seq • Spatial transcriptomics (Visium, Visium HD, Xenium) Software Engineering: Python package development • CI/CD (GitHub Actions) • Docker/Singularity • Nextflow/Snakemake • SQL (DuckDB/Postgres) • REST APIs • Rust Machine Learning: transformer & embedding models • graph-based regulatory inference • clustering & dimensionality reduction • peptide presentation scoring Translational Impact: neoantigen & peptide discovery • structural variant interpretation • biomarker identification • clinical cohort integration • rare disease genomicsProfessional ExperienceComputational scientist driving multi-omics, structural, and proteogenomic analysis pipelines across oncology and rare disease programs. Developed structural proteogenomics pipeline integrating AlphaFold, MSFragger, DIA-NN, and HLA-binding models to identify non-canonical peptides from RNA-seq, WGS, and mass spec; implemented modular Python framework with automatic QC, CLI tooling, and containerized execution. Led multi-omics analyses of CAR T cell persistence using scRNA-seq, scATAC-seq, CITE-seq, and spatial transcriptomics, uncovering cell state dynamics linked to treatment response. Architected long-read + optical mapping hybrid assembly workflows resolving structural variation in complex genomic regions (22q11.2, subtelomeres), reducing misassembled contigs by >40% and enabling population-scale discovery. Led integrative single-cell & spatial programs (Visium HD) to map tumor microenvironment remodeling in neuroblastoma treatment cohorts, identifying cell state shifts linked to therapy response. Built enterprise reproducible Nextflow pipelines for immunopeptidomics, RNA-seq, and long-read assembly used across multiple labs and HPC environments (Slurm), including automated cloud deployment. Developed full-stack analytical platforms using Flask/Django, React, FastAPI, and Shiny enabling real-time genomic interpretation and multi-omics data visualization for collaborators and clinicians. Drove independent research direction including study design, statistical modeling, computational method selection, and cross-functional communication with clinicians and wet-lab scientists.