Eric Kelsic

We developed a new technique called MAGE-seq to enable rapid and quantitative phenotyping of mutant libraries in E. coli, and apply it to understanding codon preferences in an essential translation initiation factor gene. We find that 5′ RNA structure determines codon preferences even far from the start codon and that synonymous codons that disrupt the native 5′ RNA structure are more deleterious.

We visualized evolution by recording timelapse videos of bacteria growing on a MEGA-plate (a giant agar dish) and evolving to 1000x antibiotic resistance levels over a 3 week period. We then use DNA sequencing and lineage tracing to study the mutations that confer resistance. https://www.youtube.com/watch?v=plVk4NVIUh8

We show that high-order species interactions can counter-intuitively create a lower bound on ecosystem diversity.

Guided by experimental insights, we propose a new mechanism for maintaining diversity in microbial communities that contain antibiotic producing and degrading species.

We study the structure of early Facebook collegiate social networks using community finding algorithms.