PENN STATE UNIVERSITY, College of Medicine, Hershey, PA
Research Assistant,Supervised by Dajiang Liu, PhD 2018-Present
- Literature reviews on current methods of meta-analysis and fine-mapping.
- Simulated phenotypes and genotypes for multiple ancestries under Mendelian traits scenario.
- Journal Club.
Teaching Assistant, Supervised by Jiangang Liao, PhD 2019-Present
- TA for graduate level: PHS 521: Applied Biostatistics.
- Grading bi-weekly assignments and answering questions.
DUKE CENTER FOR STATISTICAL GENTICS AND GENOMICS, School of Medicine, Durham, NC 2017-2018
Graduate Research,Supervised by Andrew Allen, PhD & Timothy Reddy, PhD
Differential Expression Genes Analysis on STARR-Seq Data Set
- Designed reasonable loci-sample specified offset for Negative Binomial Regression Model on RNA-Seq data.
- Built GLM models to identify significant interaction between allelic group and time point.
- Performed analysis using simulated STARR-seq data.
- Published a poster presentation in ENDO 2018.
DUKE UNIVERSITY, School of Medicine, Durham, NC 2017
Internship, Supervised by Megan Neely, PhD
Understanding Post-Transplant Lymphoproliferative Disorder in the ISHLT registry.
- Constructed Statistical Analysis Plan (SAP) to help communicate between investigators and statisticians
- Performed KM (Kaplan-Meier) Plot to compare survival rate between two groups on sample data set
- Applied Cox Model to detect potential confounders in the study on sample data set
- Used automatic report in SAS to construct final results on sample data set
PEKING UNIVERSITY, School of Mathematical Sciences, Beijing, China 2014-2016
Research Assistant, Surprised by Junshan Shen, PhD
Project 1: The Application of Linear Mixed Model on Data of Menopause(2015-2016)
- Built a model to determine which of three drugs were effective in treating Menopause
- Applied a linear mixed model and discriminant analysis with R and SAS to compare 3 different drugs
- Advanced missing value imputations process to handle data input/wrangling
- Created a reasonable placebo group to be used in a clinical trial
Project 2: The Cause of Uterine Endometriosis (EM)(2014)
- Built a model to determine which of different diseases are related to endometriosis(EM)
- Implemented variable selection via clustering analysis under various criterions to filter out irrelevant signals in patient test results
- Applied logistic model to establish quantitative relationship between possible causes and EM