SHAP independently recovered delayed circadian phase as the dominant depression feature from wrist-worn actigraphy, without clinical annotations. The engineering recovered the biology.
Arash Rahmani · M.Sc. Bioinformatics, Julius-Maximilians-Universität Würzburg
Repositioning from RNA-seq pipelines to computational psychiatry and digital health. The engineering philosophy that produces reproducible genomics workflows transfers cleanly to behavioral data: config-driven, schema-validated, participant-aware, pytest-tested. The data changes; the rigor does not.
Classification of psychiatric conditions from wrist-worn motor activity (n = 76, inpatient cohorts). A double dissociation: circadian timing separates disorders, activity volume separates patients from controls.
Macro-F1 0.808 binary · 0.691 3-class · paired gain +0.109 over distributional features alone (20/20 repetitions, committed fold fixtures). Preprint in preparation.
Python · scikit-learn · XGBoost · SHAP · pytest · 160 tests passing
Reproducible RNA-seq differential expression and pathway enrichment in Python. Validated on Fagus sylvatica transcriptomics under varying CO₂ conditions. Config-driven, src-layout, modular.
Python · PyDESeq2 · gseapy · pandas · NumPy
Open to computational psychiatry, digital phenotyping, and digital health ML roles. Germany · EU · UK · Switzerland