Michael Cheng Email & Phone Number

United States

Los Angeles, Ca, Us

Ucla

Development of single cell multiomic gene regulatory network (GRN) methods • Integrated spatial transcriptomics and scRNAseq in a gradient boosting-based cell-cell communication GRN inference method using spatially resolved scRNAseq data• Benchmarking inter- cell type GRNs using spatial mutual information, and ligand-receptor enrichment• Hyperparameter tuning of the lab’s intra-cell type GRN method, SCING, using Perturb-seq gene knockout experimentsDevelopment of scalable public novel cell type GRN web server and database.• Created 1000+ cell type networks across 90 tissues in 8 human and mouse cell atlases and annotated them with disease relevant modules.• Developed and maintained a network generation and analysis pipeline with parallel cluster computing.• Building a public web server and implementing a computationally efficient graph database to host all networks and analytical tools to facilitate single cell research.Elucidation of pathways involved in TGFβ induction of LRRC15 in immunotherapy resistance in late-stage cancers.• Identified cancer cell line GRNs associated with TGFβ induction of LRRC15 which is responsible for immunotherapeutic resistance.• Identified four gene targets which were experimentally validated to modulate LRRC15 expression, and clinically validated to correlate with LRRC15 and worsen cancer progression.Systems genetics study of cognitive outcomes of mild traumatic brain injury (mTBI) in pediatric patients.• Carried out GWAS analysis to identify genetic variants that impact learning and memory differently in mTBI and orthopedic injury adolescent patients.• Uncovered genetic mechanisms modulating mTBI effects on learning and memory through pathways including mitochondrial organization, neuronal development, synaptic signaling, and glial cell activation.• Developed a polygenic risk score that significantly predicts learning and memory outcomes better than existing prognostic models for pediatric mTBI patients.

Ucla

• Tools used: R and Bioconductor, Bash, Python, Neural Network SGE Computing Cluster• Performed Nonnegative Matrix Factorization (NMF) on patients from 23 cancer types in The Cancer Genome Atlas to extract unique epigenetic-factor gene programs associated with cancer survival outcome.• Trained a pan-cancer artificial neural network to accurately predict patient survival outcome from the NMF gene signatures.• First author publication on these findings featured by UCLA Health and the National Cancer Institute

Ucla

• Tools used: R, Bash, Latex, GCTA, PLINK, SGE Computing Cluster• Developed Polygenic Risk Score pipelines and data visualizations to compare prediction accuracy between GWAS summary statistics and Genetic Relatedness methods (GRM) with GCTA-simulated phenotypes and genotypes drawn from 1000 Genomes data. Optimized methods to improve early disease diagnosis and equity in healthcare• Improved GWAS SNP effect estimate significance by replacing phenotype with BLUP genetic effect from GRM

• Tools used: R and Bioconductor, Bash, Python, SGE Computing Cluster• Compared gene expression of prostate cancer cell lines and cancer cell types against normal tissues through data visualizations from RNA sequencing and alternative exon splicing data to inform researchers for their future analyses.• Identified key pathways and transcription factors in late stage cancers using Gene Set Enrichment and Gene Regulatory Network Analyses• Efficiently screened T-cell receptor candidates for further immunotherapeutic study in the lab by implementing a clustering network script based on VDJ region similarity

Doctor Of Philosophy - Phd, Bioinformatics, Gpa: 4.0/4.0

Bachelor Of Science - Bs, Molecular, Cell, And Developmental Biology With Specialization In Computing, Bioinformatics, 3.919

Activities and Societies: Marching Band, Road To Damascus A Cappella Group, Grace On Campus

Education record

Activities and Societies: Marching Band/Drum Line, Wind Ensemble, Chamber Choir, Science For Youth, FISH Club