I build computational and statistical methods for human genomics — and I care that they work for the populations genomics has left out.
My doctoral research investigates how polymorphic natural killer cell receptors shape multiple sclerosis. The KIR genes are among the most variable in the genome and their structural complexity has kept them out of reach of genome-wide studies; I wrote PONG 2.0 to recover KIR alleles from the SNP data biobanks already hold.
I trained first as a statistician (B.Sc., M.Sc.) and taught probability and statistics for six years.
Publications
-
PONG 2.0: allele imputation for the killer cell immunoglobulin-like receptorsFirst author
Human Molecular Genetics · 2026 · 35(17):ddag075 · doi:10.1093/hmg/ddag075 · PMID 42607289
-
Uncovering hidden gene–trait patterns through biclustering analysis of the UK Biobank
bioRxiv · 2024 · 2024.11.08.622657 · PMC11601405
-
Archaic HLA class I receptor modulates natural killer cell driven immunity throughout Oceania
Cell · 2024
-
SysBiolPGWAS: simplifying post-GWAS analysis through the use of computational technologies and integration of diverse omics datasets
Bioinformatics · 2023 · 39(1):btac791 · doi:10.1093/bioinformatics/btac791
-
Polygenic risk score in African populations: progress and challenges
F1000Research · 2023 · 11:175 · doi:10.12688/f1000research.76218.2
-
A review of evolutionary trends in cloud computing and applications to the healthcare ecosystem
Applied Computational Intelligence and Soft Computing · 2021 · doi:10.1155/2021/1843671
Full list on Google Scholar.
Projects
-
PONG 2.0
Research software · 2026
Imputes high-resolution KIR genotypes from SNP-array data using pre-trained multi-ancestry models. 92–99% accuracy, validated against targeted sequencing. Open source, GPL-3.0.
- R
- C++ / Rcpp
- Shell
- PLINK2
-
kir-phla-fusion
Research code · 2026
Deep learning pipeline predicting KIR-HLA-peptide binding strength. ESM-2 embeddings with concatenation fusion to preserve receptor and ligand chain identity, trained on 3,321 binding measurements curated from primary tabulated sources. Beats both the prior published model's averaging approach and a biochemical baseline under leave-one-receptor-out cross-validation.
- PyTorch
- ESM-2
- HuggingFace
- Python
- SLURM
-
SysBiolPGWAS
Research platform · 2020–2021
Deployed platform for post-GWAS analysis across diverse omics datasets: annotation, imputation, eQTL colocalization, TWAS. Serves the African genomics community.
- TypeScript
- Python
- Shell
-
Biclustering TWAS
Research code · contributor
Biclustering of transcriptome-wide association results from PhenomeXcan, grouping genes and traits together to surface overlapping patterns that global clustering misses.
- Python
- Jupyter
- pytask
- conda-lock
- TWAS
-
Pseudobulk simulation for deconvolution
Research code · 2023
Preprocessing and pseudobulk simulation pipelines supporting evaluation of BuDDI, a variational autoencoder for bulk RNA-seq deconvolution. NNLS residual analysis to find missing cell types.
- Python
- scanpy
- scikit-learn
- SciPy
- HPC
-
Computer-Based Test Platform
Desktop application · 2021
Examination system with a candidate client for timed exams and an admin application for monitoring sessions in real time.
- Python
- PyQt5
- SQL
- JavaScript
Experience
-
Graduate Research Assistant — Computational Genomics
Norman Lab, Department of Biomedical Informatics — University of Colorado Anschutz
August 2022 – Present
Lead investigator on the role of polymorphic KIR in multiple sclerosis.
- Wrote PONG 2.0 — an R package with C++ acceleration. 92–99% accuracy across five ancestry groups, validated in 267 independent samples, benchmarked in over 8,000 individuals.
- Testing KIR–HLA receptor–ligand combinations for association with multiple sclerosis in the NIH All of Us Research Program (Controlled Tier CDR v9), pairing existing HLA genotypes with PONG 2.0-imputed KIR genotypes across an ancestrally diverse cohort.
- Modelled gene–environment interactions on polygenic risk scores in REGARDS (30,239 participants, multi-ancestry).
- Built simulation pipelines to evaluate deep generative models for RNA-seq deconvolution.
- Mentor scholars in PATHways in Genomic Data Science (PATH-GDS) and PATH-GREU — NIH-funded programs training MS statistics students for genomic data science careers.
-
Teaching Assistant
Office of Research Education — University of Colorado Anschutz
Fall 2024, Fall 2025
Instructional materials, coursework assessment, and review sessions for graduate students.
-
Research Assistant
Covenant Applied Informatics and Communication Africa Centre of Excellence (CApIC-ACE), Nigeria
January 2020 – September 2021
- Co-developed SysBiolPGWAS, an open-source post-GWAS platform for the African genomics community.
- Python and Bash pipelines for fine-mapping, polygenic risk scoring, and downstream GWAS analysis.
- Led weekly lab meetings; contributed to grant writing.
-
Lecturer, Computer Science
Federal College of Agriculture Akure, Nigeria
September 2016 – August 2022
- Taught probability and statistics, and advanced engineering mathematics.
- Led a three-person team building the institution's web application and results software.
- Supervised undergraduate research projects.
Education
-
Ph.D., Human Medical Genetics and Genomics
University of Colorado Anschutz Medical Campus
Expected December 2026
Mentors: Dr. Paul Norman and Dr. Christopher Gignoux.
-
M.Sc., Statistics
University of Ilorin, Nigeria
-
B.Sc., Statistics
University of Ilorin, Nigeria
Skills
Statistical methods
Regression and interaction modelling · Bayesian inference · time-series forecasting · Monte Carlo simulation · bootstrap and permutation testing
Machine learning
Supervised prediction and ensemble methods · cross-validation and benchmarking · clustering, matrix factorization, biclustering · deep learning with PyTorch and TensorFlow · evaluation of deep generative and language models
Computational genomics
KIR and HLA imputation · GWAS and fine-mapping · polygenic scoring · population genetics across multi-ancestry cohorts · RNA-seq deconvolution
Computing & infrastructure
Python (PyTorch, TensorFlow, scikit-learn) · R · C++ via Rcpp · SQL · Bash · PLINK2 · Snakemake · Git · SLURM/HPC
Talks & Service
Conference presentations
ASHI, Orlando (2025) · European Federation for Immunogenetics, Prague (2025) · KIR Polymorphism Workshop, Colorado (2025) · ASHG, Denver (2024)
Peer review
Scientific Reports (Springer Nature), 2025 · American Society for Histocompatibility and Immunogenetics (2025) · University of Denver RaCAS (2025)
Training
ASHI Educational Workshop, Austin (2026) — ACHI-approved, 12.75 contact hours · Computational Genomics Summer Institute, UCLA (2024).
Mentoring
PATHways in Genomic Data Science (PATH-GDS) and PATH-GREU, University of Colorado · Clear Direction Mentoring — career development for STEM professionals.
Memberships
ASHG · British Society for Immunology · Society for Immune Polymorphism · ACTRIMS · H3Africa