I have a Ph.D. in computer science from University of Massachusetts Boston under the supervision of Dr. Nurit Haspel. I finished my Computational Science, Bioinformatics Co-Op at Moderna under the supervision of Christopher Pepin.
My areas of expertise and research intrests are Bioinformatics, Computational Biology, Machine Learning, and Neural Networks.
In this project I worked with three chemical probing software tools and created a Typer CLI tool and a NextFlow (DolphinNext) pipeline for running different parts of each software and comparing them
In this project I analyzed end-seq assay NGS data and the predicted RNA structures to find RNA degradation signals
In this chapter we survey both experimental and in silico methods used to study protein dynamics and explore the conformational space of proteins. We provide detailed discussions of the challenges of using these methods, their strengths and shortcomings, and their improved variants. We then focus on new machine learning-based strategies that have been a research hotspot for computational biologists in recent years.
We provide an extensive evaluation of our previously introduced, robotics inspired conformational search algorithm (RRT* with Monte Carlo). We then identify what intermediate conformations appear the most in our generated conformational pathways using TDA Mapper, a topological data analysis algorithm, and examine how close these intermediate conformations are to existing experimental data.
We developed an AutoEncoder network to create protein family fingerprints, a Command Line Interface tool to find similarity between different families using their fingerprints and to find the closest family for a new protein
We developed a Convolutional Neural Network to predict the protein-protein interaction contact map using the fingerprints of the proteins.
In this work, we integrated the rigidity analysis of proteins into our algorithm to guide the search to explore flexible regions. We demonstrate that rigidity analysis dramatically reduces the run time and accelerates convergence.
The objective of this analysis is to combine text annotation, social network analysis, and statistical modeling to identify advice-seekers on online social media for buprenorphine-naloxone use and study their characteristics.
We developed a data acquisition system, implemented signal processing techniques and machine learning algorithms to detect the picks for a mass spectrometer. I also designed an interactive UI with ability to tune the parameters of the mass spectrometer hardware using R Shiny app.
In this work we compare and assess the utility of several machine learning methods and their ability to predict the effects of single and double mutations. We in silico generate mutant protein structures, and compute several rigidity metrics for each of them. We use these as features for our Support Vector Regression (SVR), Random Forest (RF), and Deep Neural Network (DNN) methods.
Implementation of an AI bot for engineering.com website
Mass housing company customers classification project.
