Connectivity Mapping
You’ve likely already read about MIT/Harvard’s project called the connectivity map. A description of this publicly accessible project was published last week in Science. The map is a first attempt in the public domain to allow researchers to discover previously unrecognized relationships between perturbants (e.g. small-molecule drugs) and gene-expression signatures on a genome-wide scale. The Science paper outlines some of the early successes of the approach, one of which–the finding that sirolimus reproduces a gene-expression profile similar to dexamethasone in dex-resistant childhood ALL cells–has immediate and potentially life-saving clinical implications. The group is proposing that its early successes serve as a clarion call for other groups to join the project to expand its information and thus its utility.
I don’t think this version of a perturbation (connectivity) map, which is a first iteration, is going to impact the state-of-the-art in functional genomic applications at pharma. Pharma’s own functional genomic “connectivity mapping” efforts are robust enough to make this early version of CMAP little more than a nice addition to a company’s existing arsenal of data. On the other hand, there is no way that any single company could match or even come close to matching the potential information capacity of an open-source public project like CMAP. Therefore, industry should support this effort, if not with data contributions, at least with money and with small-molecule drugs for testing, much as it supports the public human SNP map and HapMap projects. In the future, the CMAP effort should transform itself from an industry nice-to-have into an indespensible tool that will help to accelerate new therapeutic disoveries and simultaneously level the functional-genomics playing field within pharmaceutical discovery.
