 |
Bhavesh Borate BS: Degree, University 3500 Sutherland Ave, #E303 |
Description of Research
Mentors: Michael Langston
Genomics, Proteomics and other high-throughput technologies are producing
large amounts of interesting data. Computers and Statistical procedures are
indispensable for the handling and analysis of this vast amount of data and to
parse the signal from the noise. This, in itself, is a very challenging task.
One of the most difficult problems to be encountered in this analysis is the
identification of a threshold above which most of the data (if not all)
obtained would make or imply biological sense. We have tried to analyse
microarray data to address this “Thresholding problem”. Currently in the
analysis of microarray data, this threshold is assumed to coincide with the
correlation value of 0.85. A random threshold of this sort cannot be validated
to be applied across all sets of data. A much more dynamic threshold that
would adjust itself to the data distribution of a particular experiment should
be vital to the analysis.
Rotation Summary
Mentor: Chris Dealwis - Fall 2004
Investigation of Mouse Mbd1 protein for recognition by RNR1 reductase.
Abstract
Mentor: Jay Snoddy - Spring 2005
Development of Software Modules for the Analysis of High-Throughput Data
from a Gene Ontology (GO) Perspective.
Abstract
Mentor: Michael Langston- Summer 2005
Utilization of graph theory methods to achieve a valid cut-off threshold for
high-throughput biological data.
Abstract
Mentor: Brynn Voy - Summer 2005
Investigation of the differential expression of genes in acute high-dose (200
rad) radiated mouse testicular tissue.
Abstract