Bioinformatics Research Opportunities
Possible advisers for research experience are listed below. Faculty interested in being listed, or modifying their listing should contact Kirk Pruhs (
kirk@cs.pitt.edu).
Protein-protein interactions (PPIs) are fundamental to biological function. The increasing availability of PPI structures in the Protein Data Bank (PDB) provides a wealth of information that can be mined to better understand and predict the properties of these interactions. We aim to analyze protein structures to better classify PPI complexes as well as make predictions about biological function. Undergraduate researchers will further develop their machine learning, programming, and statistical analysis skills while learning about protein structure and function. If interested, please contact David Koes (
dkoes@pitt.edu).
We have projects in software development, method/algorithm development (particularly in drug discovery), high throughput data analysis, and wet-lab experiments. The research program in our laboratory revolves around RNA and its applications to medicine. It is well accepted that small RNAs control many important biological processes including cell development and death. It all began in late 80's, when a company, DNA Plant Technology, tried to genetically engineer a more deep purple colored petunia by introducing more of a gene responsible for that color. Curiously, instead of producing the expected deep purple color flowers, the engineered plant yielded white and variegated flowers! The mechanism of this puzzle was solved only eight years later, when the RNA interference (RNAi) phenomenon was discovered, and was subsequently recognized by the 2006 Nobel prize. Our group has also pushed the limits of convention; using computation and experiments we pioneered the discovery of a new class of RNAs that are even smaller than microRNAs, termed unusually small RNAs (usRNAs) which, is associated with learning in mammals. To date, usRNAs represent the shortest group of functionally potent RNAs discovered in animal cells. We are now developing computational and experimental tools to help us investigate three important questions: (1) Which RNAs are involved in cancer? (2) Why are the RNAs involved in cancer? (3) What are the functions of the most interesting RNAs in cancer? To address these questions, we harness the power of multiple disciplines that include biological, physical, mathematical, and computer sciences to analyze the human genome and the RNA transcripts that are processed from the human genome. If interested, please contact Bino John (
bino@pitt.edu).
The purpose of the TRanslational Informatics Applied to Drug Safety (TRIADS) lab is to conduct research at the intersection of bioinformatics, epidemiology, and comparative effectiveness to improve medication safety for older adults. Current projects include 1) advanced methods for integrating drug-drug interaction (DDI) predictions into clinical decision support tools, and 2) exploring the use ontologies and natural language processing to determine the clinical relevance of DDI predictions. Student researchers would gain skills in the use ontologies, NLP, and the Semantic Web (especially "linked data") for biomedical research. If interested, please contact Richard Boyce (
rdb20@pitt.edu)
The members of Benos' group are interested in modeling biological processes using computational and statistical methods. We are particularly interested in developing graph theoretical approaches to identify critical DNA polymorphisms in a gene network context. Students working with us will be exposed to a learning of both the computational methodology and the biology that underlines many complex phenomena and diseases. We routinely use statistical and computational (machine learning) methods in the lab to analyze high-throughput datasets. If interested, please contact Takis Benos (
benos@pitt.edu).
Virus Discovery and Functional Viral Metagenomics. If interested, please contact Jim Pipas (
pipas@pitt.edu).
Highthroughput gene expression, microRNA, SNP, NextGen Sequencing and proteomic analysis. If interested, please contact Uma Chandran (
chandranur@upmc.edu).
Computational biochemistry and structural bioinformatics. If interested, please contact Troy Wymore (
wymore@psc.edu).
Molecular biophysics. If interested, please contact Ivet Bahar (
bahar@pitt.edu).
Using comparative genomics to understand host range expansion in a human pathogen. In my lab we are interested in understanding how the human pathogen, Toxoplasma gondii, has evolved to infect an incredible array of intermediate hosts. This feature is unique to T. gondii among eukaryotic pathogens. Our central hypothesis is that genes that have driven host range expansion in this parasite are likely to be important for causing disease in humans. To identify these genes we have taken a comparative genomics approach. Whole genome sequences are available from multiple T. gondii strains, as well as a close relative of T. gondii that has a much more restricted host range. We have also sequenced another sister species of T. gondii in the lab. We are comparing these genomes in a variety of ways, and have become particularly interested recently in the role that structural genomic variation, and in particular gene expansion, has played in the evolution of these species.
We are looking for undergraduates who are motivated to tackle these newly emerging datasets using whole-genome alignment, ortholog identification, and sequence coverage analyses. Students may also be able to test the hypotheses that emerge from their bioinformatic work on the parasites that we routinely grow in the laboratory. If interested, please contact Jon Boyle (
boylej@pitt.edu).
The bacteriophage population is huge, diverse, dynamic, and old. We have established a large collection of bacteriophages known to infect a single common host and which are therefore in potential genetic communication with each other. The approximately 250 completely sequenced mycobacteriophage genomes are genetically diverse and contain mosaic architectures, and dissecting the complex relationships between these genomes and the approximately 25,000 genes we have identified presents a considerable bioinformatic challenge. Newly developed tools enable the rapid comparison of these genomes and their genetic relationships, but there are many unresolved questions awaiting bioinformatic solutions. If interested, please contact Graham Hatful (
gfh@pitt.edu).
Alternative pre-mRNA splicing. If interested, please contact Paula Grabowski (
pag4@pitt.edu).
In collaboration with the Drug Discovery Institute, I am developing new techniques for analyzing experimental cell-level data. We are essentially looking to reverse-engineer signaling pathways based on images of thousands of cells in hopes that this will lead to new chemical signatures of cancer, novel therapeutic targets and effective drug combinations. Bioinformatics students working on this project would be involved in developing a software package for general use in extracting systems biology information from high content screening (HCS) data. This software should be able to take as input generic HCS data and perform a variety of tasks, such as efficiently calculating conventional statistical measures to determine whether the data set is useful and which parameters are informative, and automatically classifying data based on its distribution. The broader goal is to develop tools that can make quantitative predictions of heterogeneous cellular responses to various stimuli. If interested, please contact Tim Lezon (
lezon@pitt.edu).
Molecular Evolution of Development. If interested, please contact Mark Rebiez (
rebiez@pitt.edu).
