PHD Computational Biology and Bioinformatics

When I first began studying Biology seriously in my first year of high school, I was at the top of my group of friends as a gamer. When I was first introduced to RNA and DNA, I set about exploring as if it was one more form of diversion. A little while later, when I began to understand that I was exploring life itself, I became more serious and redoubled my efforts, making the study of Biology my own, even in my free time. By the time I finished high school and began studying Biology full time at Sichuan University, I no longer sought diversion, only a progressively deeper immersion into all the life sciences, especially Biology. 

Graduate School in Biology at Sichuan University

At Sichuan University I learned the fundamentals of Genetics, Biostatistics along with Cell and molecular Biology, also gaining considerable laboratory experience. Building a KRAS over-expression and NSD2 knockout zebrafish model with CRISPR and microinjection was a special highlight. one of my final projects was to perform protein docking by developing a computational approach that simulated protein-molecule interactions for modern drug design. I soon found myself fully engrossed in the study of computational methods with biological applications. Furthermore, I could not feel more certain that this area of study is my calling. 

My self-study of R programming languages resulted in an opportunity to contribute to a microarray project seeking to identify how germline variants of thePIP4K2A and GATA3 genes impact susceptibility to acute lymphoblastic leukemia (ALL). I was able to appreciate firsthand the complexity and beauty of the transcriptome.  We were able to identify several transcriptional factors binding sites located upstream of the variants and our work was published in Frontiers in Genetics. 

Nothing excites me as much as the promise of helping people with devastating diseases like ALL, using computational tools. For this reason, I could not be more resolute and dedicated to the study and application of computational biology. I earned my MS in Bioinformatics at ____ University, finishing with a solid foundation in programming, genomics, biostatistics, and machine learning, which I was keen to apply to specific biomedical research problems. Hence, I joined the laboratory of Dr. Sudeshna Das in the Department of Neurology at Massachusetts General Hospital (MGH), beginning as a co-op student and later promoted to the position of Bioinformatics Analyst.  

At MGH, I had had the excellent opportunity to delve into single-nucleus RNA sequencing (snRNA-seq) analysis for one of the largest ever postmortem brain tissue datasets. I optimized the pipeline for snRNA-seq data to identify and characterize functionally distinct astrocyte subclusters in multiple brain regions and identified key astrocyte subpopulations associated with Alzheimer's disease (AD) neuropathology and the gene pathway networks that define these cellular states. To promote open science, I built a web-based portal from scratch to make our data and other published AD datasets available to other neuroscientists. My focus on snRNA-seq data and bulk RNA-seq data analysis led me to estimate cell type-specific transcriptomic profiles from bulk RNA-seq data by borrowing information from snRNA-seq data. The difficulties lie in adjusting technical differences between snRNA-seq and bulk data, thus I am exploring various adjustment methods.  

I have been a key contributor to multiple projects utilizing various omics data modalities, including metabolomics data analysis, to reveal which ether lipid biosynthesis promotes lifespan meta-analysis of mouse transcriptomic studies. We sought to unravel context-dependent astrocyte reactions in acute CNS injury versus neurodegeneration through RNA-seq analysis of effect of APOE alleles on the glial transcriptome in normal aging and AD. ExRNA analysis in myotonic dystrophy and subcellular proteomics analysis of AD and control brains, resulted in publication. 

At MGH, I learned a great deal about methods and data for clinical research. My supervisor, Dr. ____ ____, said I have a "machine-learning mind" because I excel at "active learning" without knowledge.  Because of their black-box nature, deep neural networks are most complex. Recently, however, many interpretative tools have been proposed to reveal how deep models make decisions, especially in the field of computer vision. I am interested in using deep learning to solve biological problems through immersion in the interpretability of biological deep learning models. For instance, AlphaFold 2 (AF), launched by DeepMind in 2020, is the first computational approach capable of predicting protein structures to near experimental accuracy in most cases. Can we interpret and summarize what AF learned to improve our knowledge of protein structure? I want to explore not only how to apply deep learning to design solutions to challenging and impactful biomedical problems, but also to interpret the model, in a search for insights into the biological mechanisms underpinning the solution. 

DeepMind for discovering Biological Mysteries

I keenly look forward to honing my computational and genomic backgrounds and I am particularly excited about the opportunity to meet with the advisors that will help me to tailor the individualized elements of my curriculum to my background and interests. Thus, I hope to be selected for ____’s Computational Biology and Bioinformatics Ph.D. Program. I would be most excited to learn from and collaborate with other students as well as faculty who also are working on applying deep learning to biomedical research. Prof. especially fascinates me Mark Gerstein’s research on interpretable machine learning tools, and Prof. Kei-Hoi Cheung ‘s research on clinical natural language processing. 

Thank you for considering my application to Computational Biology and Bioinformatics at ____ University. 

PHD Computational Biology and Bioinformatics