A Web analysis suite integrates tools for analyzing drug activity, gene expression, and microRNA expression in widely used cancer cell lines.

The cancer field is awash in data capturing the molecular activity of cancer cells and their responses to anticancer compounds. But the resulting databases have become so large and complex that their information may be virtually inaccessible to many researchers trying to understand cancer and develop better treatments.

“There's been a big barrier between the people who need this information and those trained in bioinformatics who are able to access it,” says William C. Reinhold, a pharmacology researcher at the National Cancer Institute's (NCI) Center for Cancer Research. “We created a toolkit called CellMiner to help bridge that barrier and make this information readily accessible to any researcher.”

CellMiner, which is freely available at http://discover.nci.nih.gov/cellminer, integrates tools for analyzing drug activity, gene expression, and microRNA expression in the NCI-60, widely used cancer cell lines developed by the NCI for testing drug candidates.

The suite of Web tools features a pattern comparison tool that identifies statistically significant correlations between gene expression and drug activity profiles, or other patterns of interest, and that also allows input from individual experiments. “This used to be a long, cumbersome process,” says Reinhold, who is first author on a Cancer Researchpaper describing CellMiner. “Now it happens automatically. All you need to know is what you want to compare.”

A physician, for example, can plug in a drug's name and within about 15 minutes discover which genes and microRNAs track with the drug's activity across the NCI-60, and identify other drugs with similar patterns of expression or activity.

One prime application will be comparing drugs and genetic targets to identify compounds that could be effective against different forms of cancer. In an example cited in the paper, the researchers looked at colon cancer patterns and found a new compound that potentially may show greater anticancer activity than 3 compounds in clinical trials.

CellMiner currently includes data from 22,379 genes and 360 microRNAs catalogued in the NCI-60 and from 20,503 previously analyzed chemical compounds (with data from the Developmental Therapeutics Program), including 102 U.S. Food and Drug Administration–approved drugs.

The NCI team is now completing additional database analyses, and will soon add comparative genomic hybridization and whole-exome sequencing databases and tools.