0 `mo` gene symbols match between the datasets and the BLAST graph.

[houruiyan]

Hello, thank you very much for your kind help.

I meet another question again. Now I can smoothly run the SAMP(). However, the result cannot meet my exception. I always get the 0 gene symbols match between the datasets and the BLAST graph.

The following is my code

Hope to get yoru answer. Thank you very much!

[atarashansky]

Hi @houruiyan ! Notice how in the transcriptomes, you have the isoforms! (ENSG.....1234.1), whereas in your data you do not have the isoform. So the strings don't match up.

What I would recommend is to read the mapping tables (e.g. mo_to_ra.txt) and delete the .1/.2/.3s at the end of all the gene names, then save the tables. That should fix your problem.

Best, Alec

[houruiyan]

Thank you very much. Alec! Maybe it is not correct to just delete .1 .2 .3 ? The different isoform has different value in that mapping table.　So which isoform's value should I select for the certain gene? Hope to hear you. Thank you!

[atarashansky]

It's okay to delete .1 .2 .3 as SAMap will just collapse all isoforms into one node for that transcript and combine all the different mapping values.

So if A_GENE1.1 maps to B_GENE1 and A_GENE1.2 maps to B_GENE2, then A_GENE1 will map to B_GENE1 and B_GENE2.

Combining the isoforms is correct because that's what the read mapping is effectively doing if the transcriptome/gtf that you used for generating the expression matrix does not delineate between isoforms.

[atarashansky]

Closing for now! Please reopen if you're still having trouble.