Labor scietificus
Gene-Environment Interactions
With an advent of modern genotyping technologies, epidemiologists have been increasingly interested in identifying genetically defined subgroups within a population with unusual resistance or susceptibility to environmental exposures. Our work is aimed at the question of modeling gene-environment interactions for population-based case-control studies that possibly include:- environmental exposures such as dietary intake are subject to massive errors in measurement;
- missing genetic information;
- haplotype-phase ambiguity;
- genotype is subject to misclassification.
Pdf file, Web Appendix
Software , Presentation Slides
Lobach, I.V., Mallick, B., Carroll, R.J. Bayesian Semiparametric Modeling of Case-Control Data of Gene- Environment Interactions With Missing Genetic Data and Covariate Measurement Error.
Lobach, I. V. and Fan, R. Genotype-Based Association Mapping of Complex Diseases: a Case-Control Approach with Multiple Markers and Measurement Errors in Environmental Exposures.
Lobach, I. V., Chatterjee, N. and Carroll, R.J. Case-control studies of gene-gene interactions with missing genetic data and haplotype-phase ambiguity.
Population Genetics
With the availability of large numbers of genetic markers in the human genome and the advances in genotyping technology, it is becoming feasible to genotype thousands of markers on a number of individuals from multiple populations. The available data consist of a set of unphased genotypes, for each individual, and it is of interest to investigate haplotype structure since this leads to a better understanding of linkage disequilibrium patterns in the human genome and the relationship between various human populations. We are working on methodology for haplotype estimation based on a coalescence-guided hierarchical Bayes model in multiple populations.Lobach, I.V., Zhang, Y., Liu, J.S. and Zhao H. Coalescence-guided Hierarchical Bayesian Modeling of Haplotype Frequencies in Multiple Populations.
Genome-Wide Association Studies
A central problem in genetic epidemiology is to identify and rank genes involved in a disease. Recent development of high-throughput technologies generated a variety of different sources of data that can be used for detecting disease-gene associations. Genome-wide scans are increasingly utilized for analysis of massive amounts of single nucleotide polymorphisms (SNPs) in the human genome for associations with a given disease. This type of data poses challenges in statistical analysis. One is that the data is large scale. Further, complex diseases are caused by several genetic variations of small effect. Hence it is crucial to develop computationally efficient statistical methodology that allows (1) incorporation of prior biological information and (2) integration of various types of data.Lobach, I. V. and Zhao, H. Pathway-based Analysis of Large Scale Association Studies.
Lobach, I. V. and Zhao, H. Meta Analysis of Lod-score Functions for Marker Prioritization Based on Genome-wide Scans.
Home
Research
Teaching
Personal
Contact