As a junior statistician,
I am fortunate to work closely with DATA
As a junior statistician,
I am fortunate to work closely with DATA
——Three Principles of Data Science Prof. Bin Yu, UC Berkeley
Conway's Venn Diagram
Large-scale testing and Empirical Bayes
A Methodology and Computational tool for large-scale hypothesis testing
An Empirical Bayes (EB) implementation for local false discovery rate computation
Achieved better power and reproducibility by variance mixing and shape constraint
Available through Bioinformatics and CRAN package (with Prof. Newton )
Figure from Zheng et.al., 2021, Bioinformatics
Dimension constraint and latent clustering
Consider the graph/clustering information when doing million-units testing
Automatically detect the latent clustering without the adjacency matrix as a input
Simulation-based local false discovery rate (lfdr) computation
A simulation result for latent clustering inference
High density peptide array bioinformatics
A collaboration work supervised by Prof. Shelef in Department of Medicine
Design a high density peptide array containing more than 6 million peptides
Particularly interested in biomarker identification for auto-immune diseases, such as Rheumatoid Arthritis (RA), lupus, Sjogren's syndrome, etc.
Figure from Zheng et.al., 2020, Arthritis & Rheumatology
Spillover causal inference and experiment design
A time series causal inference model for online platform AB testing
A general methodology and computational tool for spillover effect
Available through CRAN package (with Feiyu Yue, Kuaishou data scientist)
A simple example of spillover experiment design
PrSS calculation and Clinical trial
An Empirical Bayes (EB) tool for Probability of Study Success (PrSS) evaluation with correlated endpoints
A research project leading by statistician group (Dr. Liu, Lin and Zhang) of Eli Lilly and Company
A generation and evaluation process for PrSS calculation