Doctor of Philosophy in Quantitative Biosciences

in the School of Biological Sciences

Nolan English

Defends his thesis:

Using heuristic derived features in machine learning to recommend post translational modifications for experimental study

Wednesday, April 13, 2022
2:30pm Eastern Time

Via BlueJeans: https://bluejeans.com/4537128340
Open to the Community

Advisor:

Dr. Matthew Torres

School of Biological Sciences

Georgia Institute of Technology

Committee Members:
Dr. Christopher Rozell; School of Electrical and Computer Engineering, Georgia Tech

Dr. Peng Qiu; Department of Biomedical Engineering, Georgia Tech & Emory
Dr. Raquel Lieberman; School of Chemistry and Biochemistry, Georgia Tech
Dr. Melissa Kemp; Department of Biomedical Engineering, Georgia Tech & Emory

Abstract:

Post-translational modifications (PTMs) alter the chemistry of amino acid residues within translated proteins and thereby have the potential to expand the function and complexity of the proteome beyond the limits of the genome. Since the advent of high-throughput protein sequencing by mass spectrometry, hundreds of different types of PTMs have been discovered enabling cell signaling, protein degradation, DNA regulation, and nearly every other cellular function. However, the rate at which PTM data are generated far surpasses the rate at which it is being curated and/or processed for interpretation. Today, more than 2 million PTMs contributing to over 400 types of modifications exist within the public domain, however far fewer PTMs are thought to have a function. Even less have their functional context understood due to the high burden of experimental evidence needed to uncover functionality. In this seminar I will describe the development of SAPH-ire, a machine learning model meant to recommend potentially functional PTMs for experimental investigation. I will present this model as part of a general data, analytics, and visualization approach meant to close the gap between PTM detection and study.