Research

My lab's research is focused on three related areas:

• Predicting functional network organization of the cell
• Network-scale analysis of complex multigenic disorders
• Decoding the genotype-phenotype map for Autism

In the first research area, we have been investigating the utility of phylogenetic profiles for predicting gene function. Phylogenetic profiles are commonly employed in gene annotation efforts, but there has been little attention paid to how the size and composition of the profiles alter their predictive power. To address this question we developed a large database of profiles and an optimization algorithm that permutes the profiles to alter size and composition while maximizing predictive power. An exciting outcome has been a small set of genomes that predict function with very high accuracy and that should make a solid core for future efforts in understanding the functional network organization of the cell.

Within the second research area we have been designing novel algorithms to identify dysregulated biological processes in disease through analysis of high throughput genomic data, including transcriptional profiles from mixed cell populations and GWAS studies. To give an example, we have worked out a new classification approach for analysis of time-series microarray data that can find processes under significant transcriptional regulation in heterogeneous tissue samples. We are now integrating these results with protein interaction networks for improved identification of significantly regulated processes in expression data in neurological disorders and other diseases.

In the third area, my lab is comparing what is known about the genetics of autism with the genetic systems of other behaviorally-related neurological disorders. One basic hope is that we will find “usual suspects” that have significant implications for neurological malfunction. A grander hope is that the work will result in a clearer genotype-phenotype map for autism, i.e. that it will enable us to circumscribe sections of the genetic landscape of autism that cause epilepsy, seizure disorder, schizophrenia, etc., thereby leading to a set of genetic markers that can be used for diagnosis/prognosis. We are also hoping to level the playing field to enable all autism researchers to tap into the benefits of computational systems biology for deciphering the genetic map of autism – by making our informatics approaches and results accessible through a web resource called Autworks