Title: Text mining for molecular network-based candidate gene
prediction

Speaker: Michael Krauthammer, Yale (Department of Pathology)

When/where: Tuesday October 12th, 4:15PM, AKW 200

Abstract:
There is an increasing interest in integrating literature information
on molecular-level processes (such as published research results on
molecular interactions between genes, proteins and other entities)
with other types/sources of information (such as genetic linkage
information). Given the ever-growing size of the biomedical
literature, much research has been devoted to automate the parsing and
extraction of research results from scientific articles. I will first
discuss our experience with working on an automated text mining
pipeline for extracting molecular network information from the
literature. In a second part, I will be presenting our research on
network-based data modeling and analysis: Using Bayesian inference
models and Markov Chain Monte Carlo (MCMC) techniques to explore
properties of molecular network information, and developing graph-
based approaches for pinpointing (genetic linkage-derived) disease
genes in molecular networks.