About Xpressomics

Xpressomics is a gene expression search engine. It identifies treatments, which your genes of interest respond to. Furthermore, it provides a quick overview of statistically significant experimental results, most of which have never been published. The search engine is based on differential gene expression estimates from thousands of public data sets including GEO, Array Express, Japanese Toxicogenomic Project, Connectivity Map and DrugMatrix.

Why are we doing this?

The problem today is that it is difficult to find relevant information about genes from already conducted experiments. Usual PubMed or Google searches reveal only a handful of publications where the genes of interest have been mentioned. However, there are vast amounts of information hidden in .CEL files from large-scale gene expression profiling studies that sit unanalyzed in public repositories. We are solving this problem. We unlock this hidden information through careful manual annotation of experiments and detailed differential expression analysis to bring you relevant information from mountains of publicly available experimental data.

Benefits for you

By searching through previously conducted gene expression experiments you can discover information about specific drugs, conditions or triggers, which induce or repress your genes of interest in a statistically significant manner. This will help you to hypothesize about the possible regulatory mechanisms and the functional significance of the genes you are studying. It also reduces the risk of failing in your next experiment as you can verify whether similar conditions have had an impact on the genes under study. Most importantly, it allows you to save time and resources by focusing on most relevant hypotheses, which have not been studied in sufficient detail.

How do we do it?

We hire research experts to manually annotate experimental factors in each experiment, where raw data is available. Then, the raw data (CEL files) is normalized and the treatments are compared to corresponding control groups. Differential expression is estimated using DEMI, a published and benchmarked probe-level algorithm specifically designed to achieve optimal accuracy in studies with small sample sizes (the majority of gene expression studies). Depending on the design, one experiment can yield a few or even dozens of differential expression profiles. Genes with a false discovery rate below 0.01 are indexed by the search engine. If your gene of interest has been differentially expressed in any of the studies we index, you will see links to the studies in the response to your query. And, remember, that you can simultaneously query several genes in order to identify studies where all of them are differentially expressed. We are continuously adding new profiles, so make sure to revisit our page regularly.

Types of data available

Supported species: human, mouse, rat, yeast, arabidopsis. More than 8,200 experiments and 136,000 samples. Currently, we already cover or are in the process of rolling out experiments from the following Affymetrix platforms: Human Genome U133A, Human Genome U133A 2.0, HT Human Genome U133A, Human Genome U133 Plus 2.0, Human Gene 1.0 ST, Human Exon 1.0 ST, Mouse Genome 430 2.0, Mouse Gene 1.0 ST, Rat Genome 230 2.0, Arabidopsis ATH1 Genome, Yeast Genome 2.0, Yeast Genome S98 Array.

Contact us and request your species of interest.

Learn more:

Our story in Nature blog
Interview in DNAdigest

Supported by: