Molecular biologist & published researcher → data scientist → business analytics.

I spent ~6 years at the bench — clinical diagnostic assay development (Pfizer), molecular diagnostics testing (BioReference), and molecular neuroscience research (UC Irvine) — where I lived in data: designing experiments, quantifying results, and publishing findings in four peer-reviewed Nature-family journals. I earned the IBM Data Science Professional Certificate to formalize the programming and modeling side of that work, and I now build end-to-end analyses in Python, SQL, and machine learning.

I like problems where scientific rigor meets real data: framing the question, getting the data (APIs, scraping, databases), cleaning it honestly, and letting the model — and the visualization — tell the story.

Python · pandas · NumPy · scikit-learn · statsmodels · lifelines · R · SAS · GraphPad Prism · SQL / SQLite · matplotlib · seaborn · Plotly / Dash · Folium · BeautifulSoup · Jupyter · BigQuery

• Dose-Response IC50 Analysis — 4-parameter-logistic (4PL) curve fitting to recover IC50 potencies across five compounds, with a concentration series spanning four orders of magnitude, Z'-factor assay QC, and asymmetric 95% CIs; the Prism/drc workflow from my Pfizer assay-development work, written in Python + R.
• Differential Gene-Expression (RNA-seq) — case-vs-control transcriptomics pipeline: CPM normalization → PCA/QC → per-gene testing with Benjamini-Hochberg FDR across ~2,000 genes → volcano/heatmap → ground-truth validation (empirical FDR 2.7%, log2FC recovery r ≈ 0.97).
• Experimental Design & Statistical Inference — the biostat workflow behind a controlled study: a-priori power/sample-size analysis, assumption checks, one-way ANOVA with η², Tukey HSD post-hoc, and Cohen's d with CIs — including a deliberately underpowered arm reported as absence of evidence, not evidence of absence.

• Hospital Readmission Prediction — predicting 30-day readmission for diabetic patients (UCI 130-hospitals, 100k+ encounters); scikit-learn pipelines, imbalanced-class evaluation, and threshold tuning for a realistic operating point.
• Clinical-Trial Survival Analysis — time-to-event read-out of the GBSG2 breast-cancer trial (Kaplan-Meier, log-rank, Cox PH), implemented in Python, R, and SAS to mirror publication and clinical-reporting workflows.
• Medical-Sales Churn & Revenue Forecasting — B2B account churn prediction plus Holt-Winters revenue forecasting on reproducible synthetic data, turned into a top-decile retention play.

• Predicting SpaceX Falcon 9 Landings — full pipeline: API + web scraping → SQL EDA → Folium maps → Plotly Dash → classification models with GridSearchCV.
• Stock Market Data Analysis — financial data via the yfinance API and web scraping (BeautifulSoup), cleaned and visualized.
• Weather Time-Series Forecasting — SARIMAX forecasting of hourly temperature with leakage-safe train/val/test splitting.
• Geospatial Crime Mapping — interactive Folium maps: markers, clustering, and choropleth.

BS in Neurobiology (Cum Laude, UC Irvine) · 4 peer-reviewed publications · IBM Data Science Professional Certificate · currently Director of Operations & Business Development, applying data + analytics to drive commercial results.

📫 Reach me: christina.rperrone@gmail.com