Tophat2 can't find Bowtie index files *.ebwt

Question

[52Teth]

I have run tophat2 using cmd:
tophat2 --bowtie1 -p 6 -G /project/Ref_Genome/GRCm38.gtf -N 2 -r 40 --library-type fr-unstranded -o ./ /project/Ref_Genome/GRCm38.genome.fa /project/rawdata/mouse-1.fq /project/rawdata/mouse-2.fq
but the following error occurs:
Error: Could not find Bowtie index files /project/Ref_Genome/GRCm38.genome.fa.*.ebwt)
I have checked that /project/Ref_Genome/GRCm38.genome.fa.*.ebwtl is produced.

My software infomation is here:
bowtie version 1.1.1
TopHat v2.0.13

Any suggestion?
Thanks a lot!

[52Teth]

I run my job on our cluster and there is only bowtie1 is installed. So when I run my job with tophat2, the error info is given. Now I am wonder how can I produce index file .*.ebwt. OR has the latest tophat2 solved this issule?

[Devon Ryan]

Just install it yourself then. You don't need cluster admins to install things into your home directory. Tophat2's support for bowtie2 tends to be a bit more up to date from what I've seen.

[kanika.151]

your index files should be in the same folder as the .fa files.

Answer 1

[Antonio R. Franco]

1. ebwt files need to be created with bowtie1 2. You need to call the index files by naming the prefix only. If your index files are named name_file_1.ebwt, name_file_2.ebwt and so on, then name_file (the common prefix name without the suffix) has to be used

[Devon Ryan]

The issue here is that a large fasta file was used, so ebwtl files (rather than ebwt) were produced. Tophat's support for that has been hit or miss.

[Antonio R. Franco]

Why are you using bowtie1 ?. I've used it with SOLiD data, but with Illumima almost everybody use bowtie2

[Devon Ryan]

I assume you meant to ask 52Teth that. I haven't used bowtie1 in quite a while, though it's allegedly still a bit faster for shorter reads.

Answer 2

[52Teth]

Thanks a lot. I have updated it and it really works. But I am wonder what the difference between bowtie1 and bowtie2. Does it influence my result using different version of boetie?

Thanks a lot again.

[Devon Ryan]

This was described in the bowtie2 paper.