Tissue phenotype Affy methods

This tool extracts differences between the distributions of the expression values of two groups of genes in a set of tissues based on Microarray (Affymetrix) expression data.

Functionality

Users may compare the distribution of two lists of genes of interest, lists taken from a microarray experiments after a differential expression analysis, genes from two clusters, etc. Alternatively, users may compare an unique list of genes versus the complete set of genes with expression values in the tissues selected (genes in user's list are excluded from this second list). This approach may be useful to see if your genes of interest are over/under expressed or randomly mapped in the overall distribution of the tissue.

Using Microarray data the list will be compared with the complete set of ensembl genes mapped into the array (also excluding genes from list). Users may select all combination of tissues.

Expression data

The application uses GNF (Novartis Research Foundation) Microarray data of an extended Affymetrix microarray chip [U133A], with more probes. We downloaded data from GNF. The set of experiments with this special chip include 79 human and 61 mouse tissues with mainly normal histology. So the variety, in contrast to SAGE libraries, comes from the tissue not from the histology state.

Statistics

We use the t-test method for our comparisons and always apply a multiple test correction by FDR.

Limit we set to perform a test with your data is at least 5 genes in each list. If you had an error message saying this might be your case, please check the files with the probes assigments and see if you have more than 5 genes in each list, if you don't, you may want to enrich your lists or change to a more common gene id.

Options to select

  • tissues: select any combination of tissues, each one has a set of libraries with expression values of tags that are mapped to genes.
  • normalization method: GNF provides expression values normalized by two methods: MAS5 (Affymetrix method) and gcRMA (Bioconductors method).
  • multiple probes expression handling: select the meassure you want to apply to the gene expression values when genes have more than one probe mapped (mean, greatest values, lowest value...)

Species

We provide data for human and mouse.

Files format

To submit your lists of genes make sure you provide a new line separated list of gene ids. Something like:

    ENSG00000195449
    ENSG00000191414
    ENSG00000195603
    ENSG00000191766
    ENSG00000192778
    ENSG00000192318
    ENSG00000195909
    ENSG00000195044
    ENSG00000191421
    ENSG00000190549
    ENSG00000194579
    ENSG00000193697
    ENSG00000192817
    ENSG00000189656
    ENSG00000189674
    ENSG00000190567
    ENSG00000195016