ONCOMINE: A Cancer Microarray Database and Integrated Data-Mining Platform

Abstract

DNA microarray technology has led to an explosion of oncogenomic analyses, generating a wealth of data and uncovering the complex gene expression patterns of cancer. Unfortunately, due to the lack of a unifying bioinformatic resource, the majority of these data sit stagnant and disjointed following publication, massively underutilized by the cancer research community. Here, we present ONCOMINE, a cancer microarray database and web-based data-mining platform aimed at facilitating discovery from genome-wide expression analyses. To date, ONCOMINE contains 65 gene expression datasets comprising nearly 48 million gene expression measurements form over 4700 microarray experiments. Differential expression analyses comparing most major types of cancer with respective normal tissues as well as a variety of cancer subtypes and clinical-based and pathology-based analyses are available for exploration. Data can be queried and visualized for a selected gene across all analyses or for multiple genes in a selected analysis. Furthermore, gene sets can be limited to clinically important annotations including secreted, kinase, membrane, and known gene-drug target pairs to facilitate the discovery of novel biomarkers and therapeutic targets.

Probabilistic model of the human protein-protein interaction network

“…RSU1 was originally identified in a screen for genes that could inhibit Ras-mediated signaling 29 and was later shown to inhibit anchorage-independent growth of MCF7 breast cancer cells 30 . Notably, we have observed that RSU1 is consistently downregulated in prostate cancer 3,24,26 , further suggesting a tumor suppressor role for this gene; however, to date, the mechanism by which RSU1 exerts its effects has not been elucidated. Here our model predicted, and we experimentally confirmed that RSU1 interacts with LIMS1, an adaptor protein involved in integrin signaling at focal adhesions 31 .…”

“…We identified coexpressed genes from 65 genome-wide gene expression data sets present in the Oncomine Cancer Microarray Database 3,21 . The 65 data sets consisted of more than 5,000 diverse microarray profiles representing many tissue types, differentiation states and cellular compositions, thus covering a broad spectrum of gene expression (Supplementary Table 2).…”

“…We generated protein interaction subnetworks for genes overexpressed in pancreatic adenocarcinoma, multiple myeloma and renal cell carcinoma, as defined by Oncomine 3 . In pancreatic adenocarcinoma, we found an oncogenic tyrosine kinase subnetwork involving ERBB2, MUC1, SHC1 and EPH2A and an invasion signaling subnetwork involving NET1, RhoA, RhoC and RAC (Supplementary Fig.…”

“…We began by assembling a collection of genomic and proteomic data potentially useful in predicting human protein-protein interactions that included model organism protein-protein interactions 1 , protein domain assignments 2 , gene expression measurements in human tissue samples 3 and biological function annotations 4 (Table 1). Based on previous reports, we suspected that (i) model organism interactions may suggest interactions among orthologous human proteins 5,6 , (ii) similar gene expression profiles across a panel of human tissue samples may identify interacting protein products 7,8 , (iii) protein domain pairs enriched among known human protein-protein interactions may suggest novel interactions 9 , (iv) shared functional annotations from Gene Ontology 4 may suggest physical interactions, and (v) that combining evidence from independent data sources may strongly predict protein-protein interactions [10][11][12] .…”

A class comparison method with filtering‐enhanced variable selection for high‐dimensional data sets

“…We used the breast cancer data sets included in Oncomine [16] unique FEVS-exclusive genes, most of which were identified by Oncomine as highly significantly associated with grade in at least one of the independent data sets (Table S1). …”