Same results whether or not hmmer/blast results included

[kdarragh1994]

I have run two trinity assemblies through the TransDecoder pipeline and was surprised to find that in one case the exact same number of transcripts was found whether or not I included the blast and hmmer information, and in the other case, more transcripts (~1000 more) were found when including the blast and hmmer results. Is this normal? Is there any way to check that the hmmer and blast hits were definitely used to make the final prediction?

[brianjohnhaas]

Hi, If the hmmer/pfam info and/or blast hit info was used to guide selection, then that info should be embedded into both the gff3 file and the headers of the transdecoder final peptide file. I agree that it would be peculiar for the results to be exactly the same in using vs. not using that info. Note, be sure to use the latest Transdecoder if you're using hmmsearch as opposed to hmmscan. Earlier versions were compatible with hmmscan results only, but the latest version is compatible with both - in case that's part of the issue. best, ~b

…

On Sun, Mar 19, 2023 at 11:22 AM kdarragh1994 ***@***.***> wrote: I have run two trinity assemblies through the TransDecoder pipeline and was surprised to find that in one case the exact same number of transcripts was found whether or not I included the blast and hmmer information, and in the other case, more transcripts (~1000 more) were found when including the blast and hmmer results. Is this normal? Is there any way to check that the hmmer and blast hits were definitely used to make the final prediction? — Reply to this email directly, view it on GitHub <#172>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABZRKX66DWY5XLLOZ62G5Q3W44QEXANCNFSM6AAAAAAWAFQQA4> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- -- Brian J. Haas The Broad Institute http://broadinstitute.org/~bhaas <http://broad.mit.edu/~bhaas>

[kdarragh1994]

Thank you, I think that was it, I now see the information in the gff3 file. Is it normal for there to be more transcripts retained following inclusion of pfam and blast info? I was expecting most transcripts to be removed resulting in a more reduced dataset.

[brianjohnhaas]

Sounds good. The blast and pfam mostly just adds coding sequences that otherwise didn't have sufficient coding metrics to be included based on sequence composition alone, so totally normal to expect more rather than less to be included. It's a way to better ensure potentially important sequences aren't excluded.

…

On Wed, Mar 22, 2023 at 11:17 AM kdarragh1994 ***@***.***> wrote: Thank you, I think that was it, I now see the information in the gff3 file. Is it normal for there to be more transcripts retained following inclusion of pfam and blast info? I was expecting most transcripts to be removed resulting in a more reduced dataset. — Reply to this email directly, view it on GitHub <#172 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABZRKX46HZPNVGQ6X6ERAXTW5MJZ7ANCNFSM6AAAAAAWAFQQA4> . You are receiving this because you commented.Message ID: ***@***.***>

-- -- Brian J. Haas The Broad Institute http://broadinstitute.org/~bhaas <http://broad.mit.edu/~bhaas>

[kdarragh1994]

Perfect, thank you!