Characteristics of Bacteriophage Altwerkus
Featured in Publication: Life Sciences PubMed
Altwerkus is a B1 subcluster Mycobacteriophage, a form of virus, that infects Mycobacterium Smegmatis, a form of bacteria closely related to Mycobacterium Tuberculosis.
Working under Dr. Emily Fisher and Dr. Joel Schildbach from the Biology Department at Johns Hopkins University, I discovered Altwerkus through direct plating and DNA isolation laboratory techniques. The Howard Hughes Medical Institute provided us with the opportunity to sequence its genome and annotate its DNA via the SEA-PHAGES Phage Hunting initiative.
Streamlining Analysis for Automated Neuronal Membrane Segmentation Pipelines (neuroXplorer)
Field: Connectomics (Computational Neuroscience)
The program makes it easier for a computer to detect neurons across three-dimensional high definition electron microscopy brain image stacks. This work was made possible under then mentorship of Dr. Joshua Vogelstein and Dr. William Roncal. The project is part of NeuroData's Open Connectome Project.
The visual illustration in this gallery was created by Lindsay Farrell.
Prediction of Autism by Translation and Immune Coexpressed Genes
Field: Computational Genomics
I worked under Dr. Shannon Ellis and Dr. Dan Arking to recreate statistical analyses of genetic variants (translational and inflammatory) found in blood, to diagnose children with Autism Spectrum Disorder. Part of my research involved operating high throughput clusters and big data using Maryland Advanced Research Computing Center (MARCC) and Joint High Performance Computing Exchange (JHPCE).
Analysis of Metagenomic Sequences
I worked in analyzing metagenomic data from soil samples taken from forested areas at Johns Hopkins University. The sequenced data was produced by Oxford Nanopore MinION sequencers and studied via use of the Kraken taxonomic sequence classification system from the Center of Computational Biology. This research project was made possible by Dr. Steven Salzberg.
Classification of Cancer Type Based on Cancer-Specific vs. Pan-Cancer Driver Genes Using Machine Learning
Field: Computational Oncogenomics
Sponsor: Institute of Computational Medicine
Under the mentorship of Dr. Rachel Karchin, this project aims to identify common somatic mutations and driver genes for specific cancer-types to enhance targeted therapeutics development.
We use machine learning techniques (random forest models) to predict cancer-type based on cancer-specific features (somatic mutations and driver genes). After comparing our cancer-specific statistical analysis to pan-cancer-based prediction classifiers, we show that our analysis performs better or the same with a reduced amount of features.