My main interest is in gaining an (asymptotic) understanding how phenotypes, such as human healthy diversity and maladies, are implemented at the level of genes and networks of interacting molecules.

To harvest as much information about known molecular interactions as possible, my group runs a large-scale text-mining effort aiming at analysis of a vast corpus of biomedical publications. Currently we can extract from text automatically about 500 distinct flavors of relations among biomedical entities (such as bind, activate, merystilate, and transport).

To sharpen our text-mining axes, we are actively designing related models and computational applications. Furthermore, in cooperation with our experimentally talented colleagues, we are striving to use text-mined networks to understand, interpret and refine high- or low-throughput experimental data. We are also computationally generating biological hypotheses that our generous collaborators are attempting to test experimentally.

My older passion is in developing and applying computational methods related to phylogenetics and evolutionary biology.