patjiang / patjiang.github.io

About Me

Computer Scientist/Software Engineer/Machine Learning Novice

Hello! Welcome to my page! I probably won't support this website indefinitely but It's a good slice into what I do on the day-to-day. Although my major is in CS/Math, I also have a keen interest in protein folding and protein dynamics; particularly, the idea of high-dimensional graph representations are what interest me the most. In the background, I am working with some of my friends in the Biological Physics department on creating a new heuristic embedding generation scheme for proteins that does not involve direct residue embedding. I have a good deal of experience working with and reading about different neural network architectures, and I will give a graphic below:

The % describes how I feel my general level of experience with each, as well as my knowledge thereof.

For the above Graphic, all of my experience implementing the models have been in python, largely through google colab, but with some work on the ASU Supercomputer as well.

Here is a similar graphic for language proficiency:

Education:

In progress:

• Bachelor's in Computer Science
• Master's in Computer Science
• Bachelor's in Math

Completed:

• N/A 😓

Papers/Textbooks I am reading currently:

• Protein Discovery with Discrete Walk-Jump Sampling
• Monte Carlo Methods (specifically chapter 11)
• Beyond Weisfeiler-Lehman: A Quantitative Framework for GNN Expressiveness
• Spherical Bessel Functions

Papers I have read:

• I will update later when I have free time

Current Positions

• ASU iGEM Team Lead
• Research Assistant @ BioProteanLab

What's in my Repository?

• Most of the things publicly available are either part of my efforts for this year's iGEM project (stay tuned) or work that I perform at the BioProteanLab. This year's iGEM project is focused on creating accessible tools for people with little to no machine learning experience to create de novo designed binding peptides to specified proteins of interest. Using a combination of industry standard and novel techniques, I will wrap up a unified workflow through multiple GoogleColab notebooks to create a cheap and affordable workflow. In the BioProteanLab, I created workflows for NMR analysis given multiple scans and generally a high-throughput method that can be run locally on colab or use jupyter to manage. I have contributed to some of the most used NMR packages in github, as well as worked temporarily with R packages.
• TLDR; NMR = BioProteanLab, Molecular Docking/Molecular Modelling = iGEM