Chronological Record of Research Groups:
2004-2008 Computational Biology group (Stanford University) headed by Serafim Batzoglou (AI Lab). During my time in the lab I worked on several projects. All of them are listed below, in order from the most current to the earliest. They included applying statistical and machine learning methods to various biological problems ranging from cancer classification and cell regulation to infering individuals genetic ancestry.
2003 Security Lab (Stanford University) headed by Dan Boneh. My work was focused on mounting a timing attack against Microsoft implementation of public key crypto-systems such as RSA, HD. Possibility of such attack is based on modular exponentiation algorithm that is used to facilitate decryption, specifically during SSL handshake.
2000-2001 Mobile Computing Lab (Stanford University) headed by Mary Baker. Now Mary is working for HP research. My area of work was constructing game theoretical model for Ad Hoc network collaboration. Such model would create dominant strategy for the participating node that assures their fare participation.
Projects:
Un-named Cancer Project - Goal of this project is to develope expression based cancer model for finding clinically relevant classification and promising direction for new drug discovery. This work aims to look at gene interactions in the form of binarry relationships to explain observed deviations in gene expression in cancer tissue, as compare to healthy one.
HAPAA HAPAA - inferring individual's ethnic origins based on dense SNP genotype.Goal of this project is to determine ancestral origin of each haploblock in an individual of a mixed ancestry. For that we use a non-parametric HMM model that encapsulates out knowledge of each population. Created by: Andreas Sundquist, Eugene Fratkin
MotifCut and MotifScan Finding regulatory motifs (or Transcription Factor Binding Sites) in inputs ranging from 100th to 100000 base pairs. Related problem is scanning a gene with known motif definition. We are using graph theoretical approach that enables us to find motif sites that do not conform to abstraction of Position-Specific Scoring Matrix (PSSM). You can try using out tools using the links below. Created by: Eugene Fratkin, Brian Naughton
You can also try our stand-alone application for Mac OS X, which is in its Beta version.
Copyright (C) 2006 Eugene Fratkin, Brian Naughton.
Jointly Serfim Batzoglou and Douglas Brutlag labs, Stanford University. This program is free software; This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.