hi
I am currently really confused. I recently submitted my samples for RNA SEQ to a company that not only does the actual running of the samples but also provide us some bioinformatic analysis - both gene ontology and pathway analyses.
We obtained about 440 DEGs which is a lot and the pathway analyses results they sent to us doesnt show any particular enrichment of any pathways (Q value > 0.05). So we thought maybe just that the DEGs were 'scattered' such that no particular pathway is enriched. The database that company used is KEGG but today one of my colleague input in the list of DEGs into a website called ConsensusPathDB-human (http://cpdb.molgen.mpg.de/) which not only can use KEGG but a lot other database as well. And the results that we obtained from this run showed about 10 enriched pathways showing Q value <0.05. There is a p value and a Q value so we assumed it is already corrected for the multiple pathway testing.
any advise? we are not sure which to trust now. Even for some hits we obtained from KEGG (same database as the one the company ran on) showed lower Q value than the Q value obtained from company.
sorry for long winded post. But i can upload a screenshot if need be. Please help!
Your pathway analysis results are only as good as the underlying source of pathway data. Different sources are going to have different levels of data available. As long as the analysis was done correctly you should be able to trust all resources. If consensus-PathDB worked then you can certainly use those results.
There are commercial options like Ingenuity Pathway Analysis (IPA) which are likely more comprehensive (but not complete by any means) but IPA requires an expensive license.
You could look at GeneSCF which is mentioned here often. Author of that package is active on the forum too.
In addition to the many database resources, there are countless methods for estimating enrichment, and it has been shown their results may vary tremendously. If the methods used to estimate enrichment are different, there is no surprise in them having different results.