Not a Typical Biomedical Engineer
I a sophomore at Duke studying Biomedical Engineering and Electrical & Computer Engineering. My primary research interests are stem cell engineering, neuroscience, and computational biology. I hope to engineer solutions to tackle the world's most serious diseases.
Skill
Experience
Education
My lab implements an innovative way to engineer functional podocyte cells using human iPSCs. Using them as models to study Chronic Kidney Disease (CKD), I identified important genetic pathways that may serve as treatment targets CKD. I am conducted in-depth transcriptomic analysis from this experience.
I formulated a primal-dual constraint learning problem to tackle protein pathogenicity classification. To achieve the desired outcome, I tested various prediction strategies and fine-tuned ESM2 models using Parameter-Efficient Fine-Tuning (PEFT) techniques, specifically Low-Rank Adaptation (LoRA). Our results demonstrated significant increase to current prediction benchmark and we are preapring the manuscript write-up.
I worked at the Barrett lab throughout the summer at Broad and Harvard through the Harvard Stem Cell Institute Internship. I utilized iPSC-derived neurons and progenitor cells to study Down Syndrome-specific morphology. During this time, I have learned techniques such as cell painting, ATAC-seq analysis, and stem cell culturing.
HSURV Abstract Book pg. 44I computationally simulated interactions between the cGAS protein and DNA-RNA hybrids in cancer cells. Additionally, I conducted protein structural analysis and design to generate desired mutant proteins using Chimera and Discovery Studio. Besides dry lab, I also learned to perform in vitro activity assay.
Working with the data team at Apple Greater China, I developed a Prophet-based time series prediction model to forecast 15-week Mac sales and tested a LightGBM model for predicting weekly iPhone sales in downstream stores. I performed feature engineering with Pandas. In the end of the 2 months internship, I successfully integrated the final Prophet model into the data team's prediction workflow.
I created an innovative Python workflow to identify cancer-related biomarkers across nine different cancer cohorts. In February 2023, I presented the preliminary research results at the Biophysical Society's annual conference in San Diego .
Copyright Bowen Jiang