Reference file?

[Ben Fulton]

Hi,

I'm working with some mouse data and trying to figure out how to make it work with Grit. I aligned the reads using BWA sampe but it tells me, " One of the read_type entries is set to 'auto' but a reference was not provided".

I see there is a reference file in the sample but I don't understand where it came from. Is this something I can create, or is there a file I can download for Mouse?

Thanks,

Ben Fulton

[Nathan Boley]

> I'm working with some mouse data and trying to figure out how to make it
> work with Grit. I aligned the reads using BWA sampe but it tells me, " One
> of the read_type entries is set to 'auto' but a reference was not provided".

That means that GRIT can't determine whether reverse complemented
reads originated from forward or negative stranded genes. If you're
calling GRIT from the command line, you need to either set
"--rnaseq-read-type" or "--reference". If you're using using a control
file, then you'll need to set the read_type field to forward or
backward.

> I see there is a reference file in the sample but I don't understand where
> it came from. Is this something I can create, or is there a file I can
> download for Mouse?

You could use the ensemble mouse reference
ftp://ftp.ensembl.org/pub/release-75/gtf/mus_musculus/Mus_musculus.GRCm38.75.gtf.gz
. (although I've never tried this - please tell me if you encounter
any difficulties)

Best, Nathan

[Ben Fulton]

There are several lines in that file that lack transcript_id's. I removed those and I got a little farther - the new error is "ValueError: fetch called on bamfile without index". Off to do some indexing...