GO term analysis for single, non-differential experiments

[jvi...@cub.uca.edu]

Hey Brian,

I am curious if Trinity/Trinotate has a way to plot GOterms when differential expression is not pertinent to the experiment?

I have 3 genotypes: apomict, parthenogen, and sexual.

However this experiment was not set up to be a differential study and sampling was both spatially and temporally different.

I simply want to generate tables for each genotype highlighting highly represented GOterms.

I attempted to follow the instructions for GOseq, however factor labeling is giving me a hard time. Is there a script in Trinity/Trinotate to generate such a file and would it even work with one factor (no inter-transcriptomic comparisions)?

Thank you for making my thesis possible,
- James

[Brian Haas]

Hi James,

I'll have a look later today.   I think I have code lying around to do this....  just gotta find it. ;-)

~b

--
You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsub...@googlegroups.com.
To post to this group, send email to trinityrnaseq-users@googlegroups.com.
Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

--

--
Brian J. Haas
The Broad Institute
http://broadinstitute.org/~bhaas

 

[Brian Haas]

If you pull the latest Trinotate code directly from github

    git clone https://github.com/Trinotate/Trinotate.git

you'll find a script that will count up the number of genes or transcripts mapping to each of the go-slim generic categories:

    Trinotate/util/gene_ontology/Trinotate_GO_to_SLIM.pl  Trinotate_report.xls.gene_ontology [include_feature_ids] >   go_slim_mappings.txt

Individual features could map to multiple slim categories and so be counted multiple times, so just be aware of that issue...

You can load the file into Excel and generate a pie chart, bar plot, or something...

If you include a 1 for the 2nd optional parameter for the script, it'll include the list of the transcript (or gene) identifiers that are binned in that slim category.

This does require that you generated the 'Trinotate_report.xls.gene_ontology' file as described here:

  https://github.com/trinityrnaseq/trinityrnaseq/wiki/Running-GOSeq
(filename 'go_anntotations.txt' used in the documentation)
and you must include parameter '--include_ancestral_terms' for that step.  You can do this for genes or transcripts and counts will reflect the corresponding feature type.

Let me know if it gives you any trouble...  it hasn't been used or tested extensively yet.

best,

~b

On Fri, Apr 21, 2017 at 1:18 PM, Brian Haas <bh...@broadinstitute.org> wrote:
>
> Hi James,
>
> I'll have a look later today.   I think I have code lying around to do this....  just gotta find it. ;-)
>
> ~b
>
> On Fri, Apr 21, 2017 at 1:10 PM, <jvi...@cub.uca.edu> wrote:
>>
>> Hey Brian,
>>
>> I am curious if Trinity/Trinotate has a way to plot GOterms when differential expression is not pertinent to the experiment?
>>
>> I have 3 genotypes: apomict, parthenogen, and sexual.
>>
>> However this experiment was not set up to be a differential study and sampling was both spatially and temporally different.
>>
>> I simply want to generate tables for each genotype highlighting highly represented GOterms.
>>
>> I attempted to follow the instructions for GOseq, however factor labeling is giving me a hard time. Is there a script in Trinity/Trinotate to generate such a file and would it even work with one factor (no inter-transcriptomic comparisions)?
>>
>> Thank you for making my thesis possible,
>> - James
>>
>> --
>> You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.

>> To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-u...@googlegroups.com.
>> To post to this group, send email to trinityrn...@googlegroups.com.

[Emilio Petrone]

Hi! 

I have two issues about the  Trinotate_GO_to_SLIM.pl script. The first one is about the 'Trinotate_report.xls.gene_ontology' file. I'm not sure if it's just the Trinotate.xls report with the gene ontology information for the transcripts.

However I tried the Trinotate_GO_to_SLIM.pl script and here is the second issue. I'm not sure if the file has the correct format output.

I ran the following command:

Trinotate-master/util/gene_ontologyTrinotate_GO_to_SLIM.pl Trinotate_sp1.xls go_annotations.txt > go_slim_mappings.txt

I'm attaching the file.

Thanks and best regards.

[Brian Haas]

Hi Emilio,

You'll need to extract the GO terms as described here:

https://github.com/trinityrnaseq/trinityrnaseq/wiki/Running-GOSeq

and then run that output file through the GO slim mapper.  

Be sure to do a 'git pull' on the repo to get the latest gene ontology data for this.  Note, this functionality isn't in the latest release.... it'll go into the next release.

best,

~brian

--

You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsub...@googlegroups.com.
To post to this group, send email to trinityrnaseq-users@googlegroups.com.

Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

[jvi...@cub.uca.edu]

Brian,

If I understand correctly, when you extract GOterms from Trinotate_report.xls, GOterm counts do not consider transcript abundance.

If not, is there a way to incorporate transcript abundance with GOterm counts for a single assembly, maybe using the trans_counts.TPM.not_cross_norm file? I'm thinking this would give me a better idea of relative GOterm expression.

Thank you,
James

[Brian Haas]

Hi James,

I don't think GOseq takes into account expression levels. For exploring expression intensities and GO, this might have what you want:

http://wencke.github.io/

there's weak support for it in the latest Trinity. There's some discussion on the google forum about it.

best,

~b

--

You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsub...@googlegroups.com.
To post to this group, send email to trinityrnaseq-users@googlegroups.com.

Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.

[jvi...@cub.uca.edu]

Quick question,

I wanted more relevant GOslim terms to plants so I downloaded the the plant goslim.obo and gzipped it.
I edited the Trinotate_GO_to_SLIM.pl script like so:

#my $go_slim_file = "$FindBin::Bin/../../PerlLib/obo/goslim_generic.obo.gz";
my $go_slim_file = "$FindBin::Bin/../../PerlLib/obo/goslim_plant.obo.gz";

Everything ran without an obvious error and I got terms more plant specific.

I just wanted to make sure this was a valid approach.

 Attached are the two goslim results using the origional goslim_generic.obo and the goslim_plant.obo.

-james

[Brian Haas]

Seems like a fine approach to me. I can look into supporting the different go slims that are available.

~b

--

You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-users+unsub...@googlegroups.com.
To post to this group, send email to trinityrnaseq-users@googlegroups.com.
Visit this group at https://groups.google.com/group/trinityrnaseq-users.
For more options, visit https://groups.google.com/d/optout.