bowtie index location question

[martinobarenco]

Dear all,

when I type the command:

~/where/my/data/is $ bowtie -a -v 2 hg19 -c GTTCGATGAAATCTTCTTTTTCTGTTTTCACTTGGG

the following message is returned:

Could not locate a Bowtie index corresponding to basename "hg19"
Command: bowtie -a -v 2 -c hg19 GTTCGATGAAATCTTCTTTTTCTGTTTTCACTTGGG

whereas it works fine when run from the directory where bowtie actually is. I am a bit loth copying the data where the program is, so is there a way to specify the bowtie index(es) path?

Many thanks,

Martino

[manoj123]

Hi Martino,

You can try absolute path of the index like below

~/where/my/data/is $ bowtie -a -v 2 /location/of/index/hg19 -c GTTCGATGAAATCTTCTTTTTCTGTTTTCACTTGGG

your index (.ebwt files) must be at /location/of/index/hg19.*.ebwt

[martinobarenco]

Thanks for the prompt answer, it works. It works also with the relative path (which I had tried initially), but the full path has to be specified. ie the path to the location of the executable is not enough.

[martinobarenco]

alternatively one can specify the path in the BOWTIE_INDEXES environment variable. I did that by adding the line below in the .bashrc file:

export BOWTIE_INDEXES=absolute/path/to/bowtie/indexes

then the path does not need to be repeated in the bowtie command.

(actually, this is well documented in the bowtie manual but the n00b that I am could not figure that one out)

[Fumumumu]

I am struggling almost same problem.

I wanted to divide bowtie job by chrmosome in order to make grid jobs.
In the beginnig, I made each chromesome in /var/indexes/mm9masked_chrXX/index/, respectively. Each base name was mm9masked_chrXX

Bowtie could find chr1 base name, but not other base names.
What's wrong?
Each bowtie job was like this;

bowtie --sam -C --best -f -Q /path/to/data.qual /var/indexes/mm9maked_chr2/indexes/mm9masked_chr2 /path/to/data.csfasta /path/to/data_chr2.sam

Error message was as follws.
Could not locate a Bowtie index corresponding to basename "/var/indexes/mm9maked_chr2/indexes/mm9masked_chr2

$ls -l /var/indexes/mm9maked_chr2/indexes
-rw-r--r-- 1 fumumumu staff 37334426 1 11 18:15 mm9maked_chr2.1.ebwt
-rw-r--r-- 1 fumumumu staff 52439512 1 11 18:15 mm9maked_chr2.2.ebwt
-rw-r--r-- 1 fumumumu staff 2439962 1 11 18:14 mm9maked_chr2.3.ebwt
-rw-r--r-- 1 fumumumu staff 26287529 1 11 18:14 mm9maked_chr2.4.ebwt
-rw-r--r-- 1 fumumumu staff 37334426 1 11 18:17 mm9maked_chr2.rev.1.ebwt
-rw-r--r-- 1 fumumumu staff 52439512 1 11 18:17 mm9maked_chr2.rev.2.ebwt

[fkrueger]

I am struggling almost same problem.

I wanted to divide bowtie job by chrmosome in order to make grid jobs.
In the beginnig, I made each chromesome in /var/indexes/mm9masked_chrXX/index/, respectively. Each base name was mm9masked_chrXX

Bowtie could find chr1 base name, but not other base names.
What's wrong?
Each bowtie job was like this;

bowtie --sam -C --best -f -Q /path/to/data.qual /var/indexes/mm9maked_chr2/indexes/mm9masked_chr2 /path/to/data.csfasta /path/to/data_chr2.sam

Error message was as follws.
Could not locate a Bowtie index corresponding to basename "/var/indexes/mm9maked_chr2/indexes/mm9masked_chr2

$ls -l /var/indexes/mm9maked_chr2/indexes
-rw-r--r-- 1 fumumumu staff 37334426 1 11 18:15 mm9maked_chr2.1.ebwt
-rw-r--r-- 1 fumumumu staff 52439512 1 11 18:15 mm9maked_chr2.2.ebwt
-rw-r--r-- 1 fumumumu staff 2439962 1 11 18:14 mm9maked_chr2.3.ebwt
-rw-r--r-- 1 fumumumu staff 26287529 1 11 18:14 mm9maked_chr2.4.ebwt
-rw-r--r-- 1 fumumumu staff 37334426 1 11 18:17 mm9maked_chr2.rev.1.ebwt
-rw-r--r-- 1 fumumumu staff 52439512 1 11 18:17 mm9maked_chr2.rev.2.ebwt

Apparently it is just a typo in the actual index files:

They are called mm9maked and not mm9masked!

[Fumumumu]

yaaaaaaaaaaaaah!





Thank you so much.....

[andrehorta]

Dears, i need your help!

When i typed the command: bowtie -p 4 rs4244385A

The folow message is returned:

could not locate bowtie index corresponding to basename rs4244385A

I was try to put the complete path, but not solved the problem.

This rs4244385A was generated for me, through the command bowtie-build rs4244385A.fa rs4244385A.ebwt;

Thanks

[kmcarr]

I am struggling almost same problem.

I wanted to divide bowtie job by chrmosome in order to make grid jobs.

Dividing the jobs by chromosome is not a good approach to splitting up the bowtie workload. If you want to divide the alignment across multiple nodes you should split up the reads into smaller groups and align each group of reads against the entire genome of your organism.

Imagine that you have a read which matches chr3 with one mismatch and chr9 with zero mismatches. Since you are running bowtie with the --best option you would only want the perfect alignment to chr9 returned. By your method bowtie is assessing alignments to each chromosome without consideration for possible better alignments to other chromosomes. You would then need to perform some sort of secondary processing on your multiple bowtie outputs to pick the best alignment(s) among all of the per chromosome alignments.

Split your reads into however many jobs you want run, align each group to the total genome and then merge the resulting BAM files using samtools.

[stoker]

I have reversed problem - my bowtie runs well with reference name (ex. hg19) but does not work with the full path of indexes, which I need for my local Galaxy.

This does not work:
bowtie -q -p 4 -S --sam-nohead /.../bowtie-0.12.7/indexes/test_ref /.../test_file.fastq

This works:
bowtie -q -p 4 -S --sam-nohead test_ref /.../test_file.fastq

Any suggestions how to make my full path reference working?

[giampe]

hi all,
I'm writing my problem about running of bowtie in this thread, even if I'm using bowtie to run another program named Paparazzi (Vodovar et al 2011).
the error coming out is:

[user1@netlab01:~/data_from_testuser1]$ paparazzi /home/user1/data_from_testuser1/paparazzi_new_version/paparazzi_conf_file.txt
P1B_Y_trim_clip.fastq GPGV
Creating target directories
Filtering host-derived reads
Could not locate a Bowtie index corresponding to basename "path_to_reference_genome_bowtie_library_including_library_basename"
Command: bowtie -v 0 --threads 4 --solexa-quals --quiet --un ../filtered/P1B_Y_trim_clip.fastq_unmatched.fq path_to_reference_genome_bowtie_library_including_library_basename /home/user1/data_from_testuser1/LIBRERIE_OTTOBRE/P1B/36trimmed/P1B_Y_trim_clip.fastq
Died at /home/user1/bin/paparazzi line 80, <$fh> line 1.


I'm frustated everything seems ok!

[carolW]

Hi
Although I use the right path to the index file, bowtie can't find it

/home/carolw/Desktop/bowtie-1.0.1/bowtie -x /home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index -r /home/carolw/rawSeq -S /home/carolw/rawSeq.sam

Could not locate a Bowtie index corresponding to basename " /home/carolw/rawSeq"
Command: /home/carolw/Desktop/bowtie-1.0.1/bowtie -x /home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index -r -S /home/carolw/rawSeq /home/carolw/rawSeq.sam

ls /home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index*
/home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index.1.ebwt
/home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index.2.ebwt
/home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index.3.ebwt
/home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index.4.ebwt
/home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index.rev.1.ebwt
/home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index.rev.2.ebwt

[GenoMax]

Order of options provided for bowtie are important. I am not sure if -x is needed and -r needs to be followed by your read files. -S needs to go with your sam file.

[dpryan]

The -x option is for bowtie2. You want

/home/carolw/Desktop/bowtie-1.0.1/bowtie -r -S /home/carolw/Desktop/bowtie-1.0.1/indexes/g1_bt1_index /home/carolw/rawSeq /home/carolw/rawSeq /home/carolw/rawSeq.sam

[carolW]

yes, I removed -x and it works

many thx