This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Tophat Error : Couldn't build bowtie index with err = 1

I used the following command to run tophat

tophat    -G $RPATH/ED.gtf    -o $RPATH/tophat/E1    $RPATH/bowtie/ED    ..../E1_L5.fastq, $RPATH/.../E1_L6.fastq

and I am getting the following error. If I remove GTF file from the command, it is running perfectly without any error.

[2015-01-10 16:10:40] Beginning TopHat run (v2.0.12)
-----------------------------------------------
[2015-01-10 16:10:40] Checking for Bowtie
                  Bowtie version:        2.2.2.0
[2015-01-10 16:10:40] Checking for Samtools
                Samtools version:        0.1.19.0
[2015-01-10 16:10:40] Checking for Bowtie index files (genome)..
[2015-01-10 16:10:40] Checking for reference FASTA file
        Warning: Could not find FASTA file .../bowtie/ED.fa
[2015-01-10 16:10:40] Reconstituting reference FASTA file from Bowtie index
  Executing: /util/opt/bowtie/2.2/bowtie2-inspect .../bowtie/ED > .../tophat/E1/tmp/ED.fa
[2015-01-10 16:10:41] Generating SAM header for .../bowtie/ED
[2015-01-10 16:10:42] Reading known junctions from GTF file
[2015-01-10 16:10:43] Preparing reads
         left reads: min. length=100, max. length=100, 14298526 kept reads (1494 discarded)
[2015-01-10 16:19:24] Building transcriptome data files .../tophat/E1/tmp/ED
[2015-01-10 16:19:24] Building Bowtie index from ED.fa
        [FAILED]
Error: Couldn't build bowtie index with err = 1
rna-seq bowtie tophat
• 18,755 views • link
updated by Ram Ram • written by BDK_compbio BDK_compbio

Hi,

I am facing this problem and using ssh can any body help me I am new to it

Error: Encountered internal Bowtie 2 exception (#1)
Command: bowtie2-build genome.fasta index
• 0 views • link
updated by Ram Ram • written by Azhar • Azhar

Please edit your post and add the command you used.

• 0 views • link
written by Ram Ram

Hi all,

Here is my genome

>1
CCCTAAACCCTAAACCCTAAACCCTAAACCTCTGAATCCTTAATCCCTAAATCCCTAAAT
CTTTAAATCCTACATCCATGAATCCCTAAATACCTAATTCCCTAAACCCGAAACCGGTTT
CTCTGGTTGAAAATCATTGTGTATATAATGATAATTTTATCGTTTTTATGTAATTGCTTA
TTGTTGTGTGTAGATTTTTTAAAAATATCATTTGAGGTCAATACAAATCCTATTTCTTGT
GGTTTTCTTTCCTTCACTTAGCTATGGATGGTTTATCTTCATTTGTTATATTGGATACAA
GCTTTGCTACGATCTACATTTGGGAATGTGAGTCTCTTATTGTAACCTTAGGGTTGGTTT

Here is my annotation

##gff-version 2
##bed_to_gff_converter.py

1   bed2gff At1NC000010 797 1343    0   +   .   At1NC000010;
1   bed2gff At1NC000020 2498    2816    0   -   .   At1NC000020;

but I still have the same error

Error: Couldn't build bowtie index with err = 1

Please help me to fix this. Thanks a lot!

• 1 views • link
updated by Ram Ram • written by SpreeFU • SpreeFU

3 answers

I had the same problem today using the comprehensive gene annotation GTF from GENCODE and the Illumina pre-build genome. The file names were not the problem but the chromosome names were annotated differently in the GTF and the genome (*.fa) files. The GTF annotation for chromosomes was "chr#" and the Illumina genome annotation was "#" (where # is the chromosome number).

You can fix this by either using a different genome, that has the "chr#" annotation or I just removed the 'chr' from my GTF with sed as follows:

sed 's/^chr//g' annotation.gtf > annotation.2.gtf

That worked pretty well for me. Hope this helps.

• 0 views • link
updated by Ram Ram • written by AidanQuinn • AidanQuinn

bowtie is looking for the reference FASTA file in the same directory that the GTF file is located in (and with the same name). Please check if you've addressed that requirement in your setup.

• 0 views • link
written by Ram Ram

I have ED.fasta (not .fa) and ED.gtf in the same directory, not in ../bowtie/ directory. I first used bowtie-index to create the index using this

bowtie2-build  -f ED.fasta  ../bowtie/ED
• 0 views • link
updated by Ram Ram • written by BDK_compbio BDK_compbio

Check if bowtie has an option to explicitly specify the name of the reference fasta file. Also, could you please move this to a comment on my answer?

• 0 views • link
written by Ram Ram

first,

bowtie2-build sample.fa sample

which will give you index file. keep the index files in the same directory. then, change name of your genome file (fast file) as "sample". Then, use your sample file without .fa extension. just write "sample" in tophat command. i think this works.

• 0 views • link
written by Mehmet Mehmet

Log in to answer this question.

More posts like this