Turning T-cell receptor repertoires into clinical insights one model at a time.
I'm a computational biologist finishing my PhD at Bar-Ilan University, where I build machine learning pipelines over large-scale immune-repertoire data to detect cancer from peripheral blood. My work sits at the intersection of immunomics, clinical ML, and translational oncology with a focus on breast and ovarian cancer diagnostics.
I'm actively looking for roles in diagnostics, liquid biopsy, and femtech/women's health where I can apply computational and ML expertise to meaningful biological problems.
Languages: Python, R, Bash
ML/DS: scikit-learn, XGBoost, LightGBM, atom-ml, PCA, cross-validation, feature selection
Omics & Data: High-dimensional immune-repertoire datasets (1M+ features), Pandas, NumPy, Bioconductor
Pipelines & Reproducibility: Bash scripting, reproducible research frameworks
Visualization: Matplotlib, ggplot2, plotly
Other: Git, Jupyter, R/Bioconductor, RNA-seq analysis
Machine learning over peripheral blood T-cell receptor repertoires to detect ovarian cancer non-invasively from a simple blood draw.
- Journal: Briefings in Bioinformatics (2024)
- Model: Gradient boosting
- Result: Average AUC = 0.98 on multiple splits of the data
Characterizing compositional shifts in the human T-cell receptor repertoire in ovarian cancer vs. healthy donors: a public annotated dataset for the community.
- Journal: Scientific Data (2025)
- Data: Publicly released annotated TCR dataset
Extending the immune-repertoire liquid biopsy approach to breast cancer detection from peripheral blood.
- Journal: npj Systems Biology and Applications (2025)
- Model: XGBoost
- Result: Average AUC = 0.96 on multiple splits of the data
Finishing my PhD and actively exploring opportunities in diagnostics, liquid biopsy, and women's health. Let's connect!