Help on building genome_lib_dir for STAR-Fusion (hg38)

356 views
Skip to first unread message

Job van Riet

unread,
Oct 12, 2015, 10:41:46 AM10/12/15
to STAR-Fusion
Dear all,

Could someone shed some light or guide me to a manual on how to generate a genome_lib_dir for another reference genome used in STAR-Fusion, in this case Homo sapiens hg38.
I take it I'm first supposed to run the prep_genome_lib.pl script in the FusionFilter package but unfortunately I'm stuck on the --blast_pairs parameter. Am I meant to run blastn (-outfmt 6) on all transcripts present in the accompanying .gtf file (gencode_v23) and give this as input or am I misreading this parameter?

Thanks for any help!

Kind regards,

Job van Riet

Brian Haas

unread,
Oct 12, 2015, 11:16:02 AM10/12/15
to Job van Riet, STAR-Fusion
Hi Job,

I'll add documentation for this later today and post again once it's available.  It'll show up here:


stay tuned.

best,

~brian


--
You received this message because you are subscribed to the Google Groups "STAR-Fusion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to star-fusion...@googlegroups.com.
To post to this group, send email to star-...@googlegroups.com.
Visit this group at http://groups.google.com/group/star-fusion.
To view this discussion on the web visit https://groups.google.com/d/msgid/star-fusion/eb450043-e691-4408-8bc3-6fd3cc174d3e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



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

 

Brian Haas

unread,
Oct 12, 2015, 9:43:44 PM10/12/15
to Job van Riet, STAR-Fusion
Documentation is now available:


best,

~b

Job van Riet

unread,
Oct 13, 2015, 4:53:50 AM10/13/15
to STAR-Fusion
Dear Brian,

Thanks for your quick reply and great help!
For others wanting to do the same thing using gencode annotation, remove the genes first:

grep -Pv '\tgene\t' gencode.v23.annotation.gtf > gencode.v23.annotation_transcriptsOnly.gtf

Kind regards,

Job van Riet

Op dinsdag 13 oktober 2015 03:43:44 UTC+2 schreef Brian Haas:

Brian Haas

unread,
Oct 13, 2015, 6:28:42 AM10/13/15
to Job van Riet, STAR-Fusion
You might also restrict it to the protein-coding transcripts.  Otherwise, there are more false positives to deal with.  I'll add a note about this in the documentation.

best,

~b


For more options, visit https://groups.google.com/d/optout.

Foivos Gypas

unread,
Apr 24, 2017, 2:02:06 PM4/24/17
to STAR-Fusion
Hi Brian

I have a question related to the restriction to protein-coding transcripts.
In my case i want to look on snoRNA fusions. If I keep only protein coding transcripts, this means that I will not be able to identify fusions where snornas are involved? Right?

Thank you in advance for your response.

Best
Foivos

Brian Haas

unread,
Apr 24, 2017, 3:39:26 PM4/24/17
to Foivos Gypas, STAR-Fusion
Hi Foivos,

Yes, that's right.   I honestly have no idea how well it'll work w/ snoRNAs...   we focused entirely on coding genes and lncRNAs here.  

best,

~brian


To unsubscribe from this group and stop receiving emails from it, send an email to star-fusion+unsubscribe@googlegroups.com.

To post to this group, send email to star-...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.
Reply all
Reply to author
Forward
0 new messages