This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Identification Of Tfs Using Pfam Or Plant Transcription Factor Database (Ptfd)

Hey guys,

I am a newbie in identifying TFs. I want to find transcription factors in our de novo assembled transcriptome, a non-model plant. I also read other posts in this forum, but I'm not quite sure how to do it.

I read a paper (PubMed ID: 21653784), the authors used HMM profiles already available in Pfam database or those from the domain alignments available at Plant Transcription Factor Database (PTFD) PTFD to identify TFs.

Would you mind giving me some guidance on this process?

Any replies or suggestions would be greatly appreciated.

Thank you!

next-gen

1 answer

One option would be to download the sequences of interest and use them as a BLAST database. That method could give you some quick results and help you refine your search, but it is not the most precise method available. Since this site already has domain alignments available I would:

  1. Download the alignments for the TFs of interest.
  2. Construct profile HMMs using hmmbuild for each TF.
  3. Search your translated transcriptome using HMMscan for those TFs.

The aforementioned programs are a part of HMMER, and the output of HMMscan is just plain text format which you can easily parse with a script or a few Unix commands.

Thanks SES.

Now I'm using hmmscan script from HMMER suit for TFs identification against Pfam-A database, using the code hmmscan Pfam-A.hmm Trinity.faa > outfile. Is it right?

As you mentioned, "Download the alignments for the TFs of interest". I wanted to construct a profile HMMs using hmmbuild, but hmmbuild needs alignfile format, how do you download or create such a alignfile format? For example, I downloaded plant TF sequences from PlnTFDB.

p.s. The Perl script PfamScan is also used for this purpose, the results seemed a little clearer. Which is better?

Regards, S.H.

@Izsph, if you click on one of the family names that will take you to a page with a "domain alignments" link.

@SES, I didn't notice that before. You helped me a lot. Thank you very much.

Log in to answer this question.