Conference Program

Bioinformatics in Drug Discovery: Integromic Molecular Profiling and the Miner Suite

John N. Weinstein, M.D., Ph.D.
Genomics & Bioinformatics Group, Laboratory of Molecular Pharmacology, Center for Cancer Research, NCI

With due apologies for the additional jargon, we’ve coined the term “integromics” to denote the assembly of information from multiple types of molecular profiling databases – typically from various types of microarrays. The underlying premise is that the different types of information can combine synergistically to provide insights into basic biology and medical possibilities. A case in point is the work of our research group on the panel of 60 diverse human cancer cell lines (the NCI-60) used by the National Cancer Institute to screen >100,000 chemical compounds for anticancer activity. We and our collaborators have been profiling those cells quite comprehensively at the DNA, RNA, protein, chromosomal, functional, and pharmacological levels. Those studies have included the following: transcript expression profiled using cDNA arrays, Affymetrix arrays (HU6800, U95, and U133), spotted oligo arrays, and real-time RT-PCR; protein expression profiled by 2-D gels and reverse-phase microarrays; DNA copy number assessed by BAC and oligo (Agilent, and NimbleGen) array CGH; chromosomal aberrations assessed by spectral karyotyping; DNA methylation assessed by bisulfite sequencing; mutations assessed by SNP chip and re-sequencing; sensitivity to >100,000 chemical compounds (i.e., drug candidates) by SRB assay.

Our proofs of principle for integromics in the medical context, with collaborators, include: (i) possible biomarkers for differential diagnosis of metastases of unknown origin; (ii) analyses that contributed to the clinical development of oxaliplatin for colon cancer; (iii) mdr1-inverse compounds as agents for possible treatment of drug-resistant tumors; (iv) asparagine synthetase as a therapeutic biomarker for treatment with L-asparaginase in ovarian cancer.

To aid in integrating the different types of molecular information, we have developed the Miner Suite of publicly available, widely used biostatistics and bioinformatics tools. Those widely used program packages, MedMiner, MatchMiner, CIMminer, MIMminer, AbMiner, GoMiner, High-Throughput GoMiner, and SmudgeMiner, are available at our web site, http://discover.nci.nih.gov. The web site also includes databases and a list of pertinent publications. Here, we will focus on the way in which all of those experimental and computational resources are fitting together in an integrative systems analysis. Our thanks to the many, many collaborators who have provided the basis for the enterprise described here. Supported by the NCI Center for Cancer Research.

Back to schedule