Find Genes That Correlate To A Given Gene (Database Mining, Not Data Analysis)

Question

[Pfs]

Hi,

I am interested in finding out whether there are genes that correlate with my gene of interest X. I do not have specific data to analyze, therefore this is more a database-mining question.

I looked at GEO profiles, but there are thousands of experiments with my gene of interest, and I cannot possibly look at each of them.

Any suggestion?

THANKS!

[Chris Evelo]

What do yo mean by correlate? If you would be looking at GEO you would be looking at genes that share the same expression patterns in many experiment. How would that be interesting?

Other databases will probably find genes that relate to your gene (e.g. on of them regulates the expression or activity of the other, or they interact somehow) or genes that are related to the same phenomenon.

You will first need a question before you can get any answers (that remark was meant to describe your approach, but actually it goes for your Biostar question as well).

[Russh]

maybe have a look at http://coxpresdb.jp/

[Larry_Parnell]

I agree with Chris - this question hangs on the intention of "correlate." Perhaps correlation of occurrence in PubMed abstracts is of interest. Proteomics data? And so forth - I could write a long list of genomics data that could be mined with bioinformatics tools to look for "correlation."

Answer 1

[David L.]

You could try Oncomine (Compendium of cancer transcriptome profiles with analysis engine).

Answer 2

[Pierre Lindenbaum]

you can try MadCow (a registration is required ) at http://cardioserve.nantes.inserm.fr/madtools/home/index.php

This identify genes that are coexpressed with a specified gene in multiple microarray data sets. Most existing sudies have showed that confirmation of coexpression in multiple data sets is correlated with functional relatedness. Madcow is a web tool questioning a coexpression data base with experiment filtering and several levels of significance. Results can be filtered, compared and annotated by identification of statistically over-represented Gene Ontology terms. Moreover, the user may visualize a coexpression network from the results by using the Cytoscape tool.

Answer 3

[Ian]

I have found iHOP useful in the past as you can find links with other genes in the literature. But this is a 'broad-brush' rather than similar expression etc.

Edit: You may find ATLAS useful as it does link with expression data in ArrayExpress.

[User 59]

ATLAS is indeed pretty useful for this kind of query

Answer 4

[W Langdon]

What is your gene called? http://bioinformatics.essex.ac.uk/users/wlangdon/rnanet/gene_description.html lists 24132 human exons for which GEO has reasonable Affy data. For each it gives the correlation with other exons.

Eg exon 6 of "Catechol O-methyltransferase" (Ensembl id ENSE00001297813) is correlated with exon 5 But anti-correlated with exon1 of C21orf29 "Protein TSPEAR precursor" and exon 1 of ALX4 "Homeobox protein aristaless-like 4". See http://bioinformatics.essex.ac.uk/users/wlangdon/rnanet/exon.php?1297813

The scatter plot 1 (remember to press plot) shows a negative correlation across GEO of -0.411. However this is varies between probes (see correlation heatmap).

URL: http://bioinformatics.essex.ac.uk/users/wlangdon/rnanet/correlation.html Use ENSE lookup button to translate Ensembl id to Affy probe ids. Press heatmap button.

URL: http://bioinformatics.essex.ac.uk/users/wlangdon/rnanet/scatter.html#208330_at.pm9,208818_s_at.pm10 Remember to press plot button.

Answer 5

[Joachim]

If you are interesting in mining gene-expressions, this might be a good start.

These tools (BioMart, BioConductor) allow you to wade through large amounts of data programmatically.

Answer 6

[Yogesh Pandit]

Try NextBio Public. It correlates your data with already curated data from approx. 9000 experiments (current status)

Answer 7

[W Langdon]

My appologies RNAnet host bioinformatics.essex.ac.uk hit a power problem at the weekend and this coincided with a public holiday in England. Hopefully RNAnet will be back soon. Bill

[W Langdon]

RNAnet is online. (Sorry for delay I was not in the office last week) Bill