search

MikeAxtell / ShortStack

ShortStack: Comprehensive annotation and quantification of small RNA genes
MIT License
88 stars 30 forks source link

Trouble with Bowtie #58

Closed mmageroy closed 7 years ago

mmageroy commented 7 years ago

Hi,

I am trying to use ShortStack to analysis Norway spruce miRNAs. Bowtie can't handle the Norway spruce genome (20Gb) in one file and request that the genome be split. When I split this file, then I can't input it as the genomefile. Any suggestions on how to circumvent this issue?

Thanks, Melissa

MikeAxtell commented 7 years ago

Bummer. What version of bowtie are you using? I thought that bowtie-build (at least version 1.2.0) just made a 'large' index for gigantic genomes. ( http://bowtie-bio.sourceforge.net/manual.shtml#small-and-large-indexes) .. though I have not done any testing on this that I can recall.

On Mon, Apr 24, 2017 at 8:05 AM, mmageroy notifications@github.com wrote:

Hi,

I am trying to use ShortStack to analysis Norway spruce miRNAs. Bowtie can't handle the Norway spruce genome (20Gb) in one file and request that the genome be split. When I split this file, then I can't input it as the genomefile. Any suggestions on how to circumvent this issue?

Thanks, Melissa

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/MikeAxtell/ShortStack/issues/58, or mute the thread https://github.com/notifications/unsubscribe-auth/AGiXiYA-FYhhp7sflgPIZhjMBkVI-f30ks5rzI_ugaJpZM4NGEQT .

-- Michael J. Axtell, Ph.D. Professor of Biology Penn State University http://sites.psu.edu/axtell

mmageroy commented 7 years ago

Thanks, I looked at the version on our server and it is 1.0.0. I will try it with 1.2.0 instead although in the bowtie manual they describe large genomes as human or mouse. I know that bowtie2 can handle spruce, but I understand from your description why you haven't used this version.

Best, Melissa

BioFalcon commented 7 years ago

Hello, I just experienced something similar, and found out that after bowtie generates the large index, it has to be specified as an option to use a large index on the bowtie command using --large-index, otherwise it won't work. Perhaps it could be included in future releases?

MikeAxtell commented 7 years ago

Thanks for the tip Francisco! I'll test it out on Spruce when I have a chance and incorporate fix into next release.

On Mon, Jun 19, 2017 at 6:31 PM, Francisco Falcon notifications@github.com wrote:

Hello, I just experienced something similar, and found out that after bowtie generates the large index, it has to be specified as an option to use a large index on the bowtie command using --large-index, otherwise it won't work. Perhaps it could be included in future releases?

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/MikeAxtell/ShortStack/issues/58#issuecomment-309591925, or mute the thread https://github.com/notifications/unsubscribe-auth/AGiXiTODPi83UWuoJaSuwW8TTq1gfPBkks5sFva6gaJpZM4NGEQT .

MikeAxtell commented 7 years ago

Working on this and observed one item that may be part of problem: The latest bowtie-build (1.2.1.1) no longer seems to auto-detect when a genome is 'large'. The older version, 1.2.0, did. So if I pass a large genome into bowtie-build 1.2.1.1 it crashes. It seems to need the explicit argument --large-index in order to handle indexing large genomes. I've left a ticket at the bowtie source-forge page for the developers to examine.

I am still testing other aspects of this issue for a fix.

MikeAxtell commented 7 years ago

Fixed bug with release 3.8.3. Thanks again.

mmageroy commented 7 years ago

Great! Thanks!