• Data sets
• Strain coverage
• Annotation notes
A collection of gene expression microarray data sets (strain surveys) as well as functionality such
as correlations with phenotypes and strain outlier analyses.
See also MGI's GXD database
which has expression data and images from many other types of study designs.
MPD requirements for gene expression data: strain survey of 15 or more strains; sexes segregated; 3+ biological replicates preferred.
Treatment-control and other multi-cohort studies are accessioned as multiple data sets.
• Interpretation of these data, and usage caveats
Starting with a particular gene or genomic region....
Begin by doing an MPD search on the gene symbol or location (e.g. Abcd2 or 4:62000000-63500000 or 4:62-63.5).
Proceed to the MPD summary page for the gene or region and find the gene expression section.
From there you can:
• List probesets in vicinity
• List phenotype measurements that are mathematically correlated to probesets in vicinity
• List strain outliers for probesets in vicinity
Starting with one or two strains of interest....click here
will scan the entire genome (or a selected chromosome or region) and list locations where
one or two selected strains are expression outliers relative to the other strains tested.
Analyze all data sets or just selected ones.
Starting with some strains of interest and one data set....
This tool lets you specify
two sets of up to 8 strains each and the tool will list the genes / probesets where there's
the most expression difference found between your two strain sets.
Analyze one data set at a time.
Differential expression analysis with strains of interest ...
Studies are available comparing treatments vs. controls (high-fat diet, LPS, TCE, OxPAPC) as well as females vs. males.
Starting with MPD phenotype measurements
Navigate to phenotype data of interest.
Begin by clicking on the Apply Tools link. When you get to the toolbox page click on the green gene expression icon.
This will list genes and locations that have probesets mathematically correlated to the phenotypes.
A more advanced approach is to gather the desired phenotype measurements into your shopping cart,
then go to your cart, click Apply Tools, and go from there.
• Correlate to your own strain means
Type in your unpublished strain means and find expression data (probesets and genes) that are mathematically correlated.
• Download file of all computed correlations:
All computed correlations between probesets and MPD phenotype measurements with correlation p value ≥ 0.005
This file is available by request (email@example.com)
• Downloading expression data sets: Most expression data in MPD are available via public resources such as GEO.
However, if a particular type of download file is needed let us know, we may be able to help.
• Ranked lists
• Genes correlated to the most phenotypes
• Phenotypes correlated to the most genes
• Video tutorial sessions 7 and 14 include demonstrations of MPD's gene expression functionality.
28 Mar '13 –
New data sets GX-Orozco1, 2, 3 added.
13 Mar '13 – The GX-Lusis1 data set was reloaded to include all 99 HMDP strains.
20 Dec '12 – All probeset coordinates were updated to GRCm38.
29 Mar '12 – New data set GX-Gatti1 added.
2 Mar '12 – When specific data sets are selected for the tools, multiple choice is now supported,
and the data set selection is persistent using a cookie.