How Can I Compare Differential Expression Data From Mouse And Human?

Question

[Stephanhart]

Hello,

I have run differential expression analyses for mouse and human using transcriptome data. I now have a list of genes (Ensembl IDs) differentially expressed in the mouse and a list of genes differentially expressed in the human. How can I now see if there is any overlap between the lists? Clearly I cannot use the Ensembl IDs, and many gene symbols differ between mouse and human. Any advice would be appreciated.

Thanks Stephan

[Jeannie]

Thank you very much for your answers.

Answer 1

[Neilfws]

There are various ways to map mouse IDs to human orthologs. One way is to use BioMart. Briefly:

1. Click MARTVIEW (top menu)
2. Choose database ENSEMBL GENES 64 and dataset Mus musculus genes
3. Click Filters (left menu) and expand GENE
4. Check ID list limit and choose Ensembl Gene IDs from the dropdown box
5. Either paste or upload a file of mouse Ensembl gene IDs, one per line
6. Click Attributes (left menu) and expand GENE
7. Uncheck Ensembl transcript ID (unless you want it)
8. Check "Homologs" (top of page) and expand ORTHOLOGS
9. Scroll down to Human Orthologs and check Human Ensembl Gene IDs
10. Click "Results" (menu top left)

This should return a list of mouse Ensembl gene IDs with the corresponding human Ensembl gene ID. For example:

Ensembl Gene ID    Human Ensembl Gene ID
ENSMUSG00000027168  ENSG00000007372

You can now export the results and for those differentially-expressed mouse genes with a human ortholog, find which human genes are in your human list. This may require a little more computation (e.g. a script, database or perhaps spreadsheet), to do the list comparison.

[c_u]

this is an awesome answer! I wanted mouse to rat homologues, and this did it. Thank you so much!

Answer 2

[Malachi Griffith]

You can obtain the Ensembl Orthologs for each Ensembl gene from the Ensembl Compara database using the Ensembl Compara API (or other access methods).

For example, starting with a list of human Ensembl gene IDs you can obtain the corresponding list of mouse Ensembl ortholog IDs. Once you have a map of all human Ensembl IDs to orthologous mouse IDs it should be straightforward to determine which genes in each list of differentially expressed genes are orthologues of each other. You may wish to define a minimum percent identity between orthologous sequences. Refer to the Homology Objects section of the Compara API documentation for information on how to obtain the percent identity, Ensembl IDs (aka 'stable id'), etc. for orthologs.

You will need to know the NCBI taxon ID for human (9606) and mouse (10090).

Answer 3

[swbarnes2]

A different tact; do something like pathway analysis on each dataset, and see if the same pathways turn up in each organism.

[Di Wu]

Hi, I also have similar issue that I have both blood microarray from patients and mice. I plan to combine them together. If I check the DEGs, most of them didnt overlapped. But if I separetely do enrichment analyse on the DEGs, it seem that they have similar or same functions and pathways. But I dont know how to do the further analyse, should I pick the same pathway or functions for further validation? Is there any paper for me to learn? Thank you