About me

Work Experience

Amgen, Clinical Biomarkers and Diagnostics, United States

Data scientist (R, Python, Jupyter notebook, Gitlab, SQL, Gitlab, bash command) May 2019 Oct 2021- Jun 2024

Biomarker ETL: (Python)

● Possesses a fundamental understanding of ADaM and SDTM data standards and clinical domains and Managed diverse biomarker assay types, including Blood samples, Circulating Tumor Cells, immunohistochemistry, and SLD (tumor diameter).

● Harmonized immunoassay, flow cytometry data, and lab data and integrate with patients’ infusion and adverse event timepoints, delivering over 7 indications for 13 programs.

Association pipeline & visualization for clinical data: (R, bash)

● Led the development and maintenance of the association pipeline, automating functions to generate dynamic statistical insights for analytes in biomarker clinical data, Olink, proteomics, RNA, and DNA data.

● Associated safety and efficacy endpoints with analyte data to explore data-driven univariate analysis, including baseline and various on-treatment settings.

● Delivered 400 reports across programs and provided systematic comparisons for endpoints across different indications, including blincyto, tarlatamab, BiTE.

● Created uniformed journal quality visualization for association pipeline’s scientific findings in R and Python and implemented code for quality control in regular updates.

Machine learning for adverse event predictive model: (Python, R)

● Established a pipeline for constructing predictive models capable of explaining clinical data, integrating cross-validation methods, and employing various machine learning classification techniques, including Random Forest (RF) and XGBoost.

● Conducted preprocessing of priority baseline/on-treatment biomarker features using an ensemble ranking system and managed data imputation and labeling through cross-validations. Selected features based on Lasso Regression, Elastic net, etc.

● Developed a comprehensive report integrating visualizations to highlight feature importance in various classification for performance assessment and key biomarkers. Survival analysis pipeline & visualization for clinical data: (R, bash)

● Conducted survival analysis using censored patient data as endpoints, extending findings in various analytes, including clinical biomarkers, RNA TPM data, and gene function in DNA data.

● Produced visual representations including Kaplan-Meier curve with logrank test statistics and a forest plot showing hazard ratios, and generated summarized test statistics table for the analysis.

Analysis for biomarker and diagnostic development: (R, bash)

● Collaborated with the team to optimize the biomarker analysis plan for Tezepelumab Phase 2b efficacy and safety evaluation, monitored and QC’d biomarker sample collection data routinely with the global biomarker operation lead.

● Incorporated patient stratification at baseline using inflammatory cytokines and Employed Linear Mixed Models to evaluate on-treatment effects differences between anti-IgE naive and experienced patients.

Research experience

Association analysis for Oxford Hip Score

The Oxford Hip Score (OHS) is a joint-specific, patient-reported outcome measure designed to assess disability in patients undergoing total hip replacement

Columbia University Irving Medical Center, Obstetrics & Gynecology New York, United States

(Dock, Bigdata Query, SQL,Excel Pivot,R) July 2020

● Utilized docker, R,SQL to perform ETL process for the preterm birth lead variant dataset from UK biobank.

● Expanded candidate variant using LD blocks and run variant effect prediction, Phewas analysis and pathway analysis.

● Identify the any tissue specific SNPs cluster in different enhancer and promoter types within reproductive tissue. Conduct statistical tests to detect enhancer and promoter enriched regions.

● Using HiC ATAC-seq signal data to identify preterm birth related promoter and enhancer region and focusing on finding the promoter-promter interaction

● Employing fisher exact test for enrichment analysis for eQTL data in the cohort preterm birth.

Education

Columbia University, Mailman School of Public Health, New York, NY

Master of Science in Biostatistics August 2019 – April 2021

● Computational Skills: R, Python, SAS, Sql

University of Illinois at Urbana-Champaign Champaign-Urbana, IL

Bachelor of Science in Actuarial Science and Statistic August 2015 – May 2019

● Computational Skills: R, SAS

Selected Courses

● Data Mining, Relational Database & SQL Programming, Data Science, Longitudinal Data, Statistical Inference, Biostatistical Methods I & II, The Latent-Variable Structure & Modeling

● Linear Algebra and Financial Applications, Methods of Applied Statistics, Applied Regression and Design, Applied Bayesian Analysis

Life as graduate students

Data Science project


please view my screencast for it here: