shorter time for building bowtie index

[adrian]

Hi:
I am not sure if this question has been asked before.

I have 120 RNA-Seq fasta files. I am using tophat2 for aligning to human genome.

I am queing this job on a 70 node cluster. I am using bowtie genome files downloaded from Illumina genome website.

Homo_sapiens/Ensembl/GRCh37/Sequence/Bowtie2Index/genome

I see for every fastq file alignment, tophat building index file from genes.fa..

It takes more than 30 minutes for every file.

Can this time consuming process be avoided since we already have index files for bowtie.

Following is my command line:

tophat2 -p 16 \
-G /Refs/Homo_sapiens/Ensembl/GRCh37/Annotation/Genes/genes.gtf \
-o tpht \
/Refs/Homo_sapiens/Ensembl/GRCh37/Sequence/Bowtie2Index/genome \

myfile_1.fastq myfile_2.fastq

Thanks

[blancha]

I don't know the answer to your question. Probably not.

You can however get faster TopHat runs simply by adding the option --no-novel-juncs.

[gringer]

Yes, in the most recent version(s) of tophat, you can pre-generate the indexes with the 'transcriptome-index' option:

http://ccb.jhu.edu/software/tophat/manual.shtml#t_idx

[blancha]

@gringer

Very interesting, and useful. Thanks for the information.

[adrian]

Thank you so much! that is helpful.

However when I use the following I don't see files - known.gff, known.fa, known.fa.tlst, known.fa.ver and the known.* Bowtie index files in the directory.

following is my command:

tophat2 -p 16 \
-G /Refs/Homo_sapiens/Ensembl/GRCh37/Annotation/Genes/genes.gtf \
-o tpht \
/Refs/Homo_sapiens/Ensembl/GRCh37/Sequence/Bowtie2Index/genome \
--transcriptome-index=known
myfile_1.fastq myfile_2.fastq

Is there something wrong. Where would the indexed known files will be deposited?
thanks
Adrian

[GenoMax]

To prepare transcriptome indexes you should not include the -o tpht or the sequence files. This is a special one time tophat run (Check the example in the link gringer included)

Your command would be:

$ tophat2  \
-G /Refs/Homo_sapiens/Ensembl/GRCh37/Annotation/Genes/genes.gtf \
--transcriptome-index=IF_you_want_a_directory_name/known \
/Refs/Homo_sapiens/Ensembl/GRCh37/Sequence/Bowtie2Index/genome

This needs to be a non-threaded run so you have to omit the -p 16 as well.