Our research is on the interface of complex data analysis and dynamical modeling in biological systems. We use techniques from non-linear dynamics, machine learning, and statistical physics. Most of the research has been on an intracellular scale, though recently we also developed an interest in problems at the population level. Current interests are outlined below:
Data analysis and modeling techniques from quantitative biology are used to understand COVID-19 transmission and mortality in a population. The goal is to understand long-term disease behavior if it becomes endemic, including risk factors for potential future infection outbursts. For examples, see our COVID-19 work up to now.
CRISPR/Cas is an advanced bacterial immune system, which has revolutionized biotechnology. How is this normally silent system induced? While it is hard to experimentally observe the system dynamics, this can be more readily done mathematically, where we use a combination of statistical mechanics and dynamical system modeling. CRISPR/Cas is a potentially powerful barrier to horizontal gene transfer, so modeling its regulation also helps to better understand how antibiotic resistance and virulence genes are disseminated.
Our research focus is non-canonical CRISPR/Cas functions. It is becoming increasingly clear that the system is also involved in regulation of endogenous bacterial genes that are mainly associated with bacterial pathogenicity. As the system is active under poorly characterized conditions, the target genes can hardly be systematically identified through experiments alone. We develop bioinformatics methods for predicting targets of crRNAs and CRISPR-associated small RNAs. We also computationally work on a related problem or CRISPR/Cas adaptation, specifically what sequence determinants allow distinguishing self (host) from non-self (viral) DNA.
As a model, we use restriction-modification (R-M) systems, which are often spread by horizontal gene transfer. Consequently, expression of the restriction enzyme and the methylase has to be tightly regulated during its establishment in a naive bacterial host, which is often exhibited by specialized transcription regulators. These systems come with a variety of architectures (convergent, divergent, linear), where we show that these differences are consistent with few general constraints on the gene expression regulation. Our current focus is on how these general dynamical features remain robust under changing cellular growth conditions, that affect crucial intracellular parameters.
