One of the largest challenges in genomics studies is determining the relationship between genotype and phenotype and then applying this knowledge to design principles. Metabolic engineering of bacteria can introduce targeted genomic interventions to well-characterized genes for the purpose of modifying cellular metabolism, but in some cases, even for the model organism Escherichia coli, alternative strategies are required to achieve a desired phenotype. Metabolic evolution involves applying selective pressure to a population, and over time advantageous mutations will arise that improve organism fitness. To understand what mutations occurred during these experiments and how they affect phenotype, whole genome sequencing is required, followed by mutation analysis and strain characterization.

Genome sequencing generates a large amount of data for researchers to examine and traditionally mutation analysis focuses only on gene variations. Supporting mutation analysis with computational tools and using a systems-level approach that utilizes public databases describing gene regulation and cellular metabolism improves upon existing analysis techniques and advances our understanding of how genotype relates to phenotype.

Using our mutation analysis software, E. coli Variant Analysis (EVA), we examine antibiotic resistance, benzoate tolerance, and octanoic acid tolerance in E. coli. Our analysis pipeline includes a defined set of rules for mutation categorization. Prioritization of mutations supports efforts to reverse-engineer evolved strains and focus on the variants most likely to be damaging or relevant to phenotype. From mutation analysis results, we construct biological networks for visualization of mutations and possible downstream effects. This allows for improved mutation interpretation and identification of possible mutation interactions. Furthermore, we integrate RNA-seq data into our analysis to investigate the effects of variant regulators on the transcriptome. In contrast to existing methods which focus on mutated genes, we incorporate annotations for binding sites and other regulatory features on the genome for the most complete interpretation based on the available genome and gene regulatory models.