UCSF

The genomic revolution of the last twenty years has launched the biosciences into a new frontier. For scientists working on many organisms, the availability of a genome sequence has dramatically accelerated research. The effects have been profound: Genetic engineering is now standard laboratory practice; evolutionary biologists can trace entire genomes; and genetic risk factors have been defined for many human conditions. However, our rapidly accelerating ability to collect and assemble DNA sequence information has greatly outpaced our ability to assign biological meaning to it. This imbalance creates a need for high-throughput methods aimed at developing leads to gene function: phenomic technologies. Phenomic technologies seek to associate DNA sequences with phenotypes in a high-throughput manner to gain a functional understanding of genetic code.

Bacteria are ideal candidates for phenomic analyses because of their amenability to genetic manipulation and ease at which they can be grown and analyzed under varying parameters in-vitro. In addition, even the best-studied prokaryotes like E. coli lack functional annotation for a large fraction of their genes. Therefore, bacteria present excellent experimental systems to establish the power of phenomic approaches to generate testable hypotheses of gene function en masse.

Using E. coli as proof-of-principle, we show that combining large-scale phenomics with quantitative fitness measurements provides a high quality dataset, rich in discovery. The identification of >10,000 growth phenotypes allowed us to study gene essentiality, discover leads for gene function and drug action, and understand higher-order organization of the bacterial chromosome. We highlight insights concerning a gene involved in multiple antibiotic resistance and provide information about synergy of a broadly used combinatory antibiotic therapy, trimethoprim and sulfonamides. This dataset, publicly available at http://ecoliwiki.net/tools/chemgen/, is a valuable resource for both the microbiological and bioinformatic communities as it provides high confidence associations between hundreds of annotated and uncharacterized genes as well as inferences about the mode-of-action of several poorly understood drugs. This approach can be used for all culturable microbes with available ordered mutant libraries.