Bowtie Index Files

[zun]

Hi, everyone

I am a little confused about bowtie index files.


I am treating Arabidopsis and Oryza RNA-seq data.
First, to get index file for Arabidopsis reference genome, I run this command.


bowtie-build -f TAIR9_chr_all.fas TAIR9_chr_all


It output these index files,

TAIR9_chr_all.1.ebwt
TAIR9_chr_all.2.ebwt
TAIR9_chr_all.3.ebwt
TAIR9_chr_all.4.ebwt
TAIR9_chr_all.rev.1.ebwt
TAIR9_chr_all.rev.2.ebwt


Then, I run tophat command and the result was good.

BUT, for Oryza there are still index files which were made by my collegue.
These index files are below,

IRGSPb5.1.ebwt
IRGSPb5.2.ebwt
IRGSPb5.3.ebwt
IRGSPb5.4.ebwt
IRGSPb5.fa.masked.index.amb
IRGSPb5.fa.masked.index.ann
IRGSPb5.fa.masked.index.bwt
IRGSPb5.fa.masked.index.fmv
:
:
IRGSPb5.fa.masked.index.sa
IRGSPb5.fa.masked.index.sai
IRGSPb5.rev.1.ebwt
IRGSPb5.rev.2.ebwt
# I wonder what *.index.* file is...


when running tophat with these index files, I got such warning .

Warning: junction database is empty!


At first, I doubt RNA-seq data which may be low quality.
I calcurated the both of Arabidpsis and Oryza QV , however,
their QV were almost same.


I moved *.1~4.ebwt and *.rev.1~2.ebwt files from orysa index directory
to another directory.
then, I tried Tophat again pointing new index directory.
This time I got junction database from Oryza.


Are *.index.* files not neccessary and should I not include them in the index directory?
So, what are these *.index.*?

It is earlier to ask my colleague about this,but he is now in Brazil for conference ..


I am appreciated your help!

zun