My research is focused on the usage of mathematical and computational approaches to attack biological problems in a mechanistic and meaningful way. Enhancing our knowledge by extracting information from the vast sea of biological data. 

Codon Usage Bias

The usage of synonymous codons encoding for the same amino acid differs between organisms and even within a single genome. One explanation for the heterogeneity of codon usage bias (CUB) is varying GC-content but this assumes very weak selection on CUB. Another explanation is selection so that CUB and bias in the tRNA pool match, increasing translation efficiency for highly expressed genes. The yeast Saccharomyces kluyveri is an example of an organism having a remarkable shift in GC-content: one contiguous 1 Mb region of the genome has a GC-content of 53% versus 40% in the rest of the genome (Payen et al. 2009). When ignoring this heterogeneity, existing models have to fail since the data is of contradictory nature. We hypothesized that mutational bias towards GC-rich codons is responsible for the shift in GC-bias, rather than differing selection pressures on translation efficiency. We applied our mechanistic model for CUB that estimates selection against ribosome pausing in high expression genes, expression level, and mutational biases to deconstruct the contribution of selection and mutation to the evolution of CUB of individual genes. Our results show that the whole genome experiences the same selection environment but mutation bias appears to be driving this significant intra-genomic shift. This work provides insights into the role of mutation bias on GC-content and provides a template for detecting differences in mutation and selection environments across tissues and organisms.

Spread of Invasive Species

The shipping industry is an important unintentional pathway for novel species introductions, and is responsible for some of the largest invasions in history (e.g., rats).  Understanding the patterns of species movement through shipping routes will provide vital information for preventing future invasions. Yet surprisingly few studies have attempted to model these introductions using empirical shipping networks (Kaluza et al. 2010, Seebens et al. 2013), and none have attempted to do so with terrestrial species.  Creating a mathematical model that predicts how shipping networks move both terrestrial and marine species across the globe is therefore an important task. Using shipment data from actual cargo routes, will help to visualize the system as a network, where major harbors are represented as nodes and ship movements as edges between them.  Movement along the edges is proportional to the transported cargo and the chance of a boarded individual to survive the journey. To account for further spreading this work also includes a demographic model so species can spread an establish. With this work one can identify the ports most susceptible to specific invasive species, and will be able to make predictions about how changes in trade routes will impact future invasions.