Hi Marc

I am using MapMan to analyze my Affymetrix wheat genechip data after selecting wheat mapping file provided in Mapman. I actually found that most description generated by MapMan are derived from Arabidopsis thaliana genomic database/mapping but when I blast a probe id using websites for wheat probe annotation such as PLEXdb I sometimes get different annotation with high identity%. This website provides annotation via blasting the probe id against many database including BLASTN DFCI Wheat Gene Index, BLASTN TIGR Plant Transcript Assemblies and BLASTX Arabidopsis thaliana genome. The latter gives the same annotation I found in MapMan however the identity% is always very low and the two former give high identity%. So, in my situation do I go for MapMan annotation/description or should I blast each probe id of interest to get the best and top hits of annotation? Please advise.

With many thanks
Thana Khan

Hi Thana,

it is true that, in the mercator annotation process, the arabidopsis genome has a strong impact - and hence
many of the annotations made are based on hits in the arabidopsis genome. We do, however, also include
rice as a monocot model, various mixed (concerning mono- and dicot"ness") SwissProt plant proteins and
protein domain and -family profiles from InterPro, CDD and KOG to give a better coverage over the plant
kingdom. The specific databases you mentioned in your post (PLEXdb, DFCI wheat and TIGR) are not
included in the current mercator pipeline. If you find high scoring hits in one of these DBs that have a credible
annotation (ideally even backed by experimental evidence) that differs from the annotation made by mercator this
annotation would obviously be preferable. Unfortunately that will require you to manually curate the MapMan
annotation file for the wheat affy chip and edit all the annotations that can be improved by using the DBs
you mentioned (which is a tedious process).

I hope this answers your question - Best regards,
Marc