tophat-fusion-post result sam_header empty

[onsh]

Hi all,

I’m having a problem with tophat-fusion-post. I run it, but the output is empty.
I think a sam_header file may cause that problem.

I use the version of
tophat : 2.1.0
bowtie : 1.2.2
on Windows Subsystem for Linux.

My command like this:

tophat-fusion-post -o /home/onsh/src/tophat-2.1.0/tophat_SRR064286/post_out/ \
--num-threads 16 --num-fusion-reads 1 --num-fusion-pairs 2 --num-fusion-both 5 \
/home/onsh/src/tophat-2.1.0/Homo_sapiens.GRCh38.small

Bowtie indexes are downloaded from Ensembl. In the original file,
I added a word “chr” to each chromosome numbers because the chromosome number is just a number.

My directory structure like this:

onsh@PC3:~/src/tophat-2.1.0$ ll
drwxrwxrwx 0 onsh onsh 512 May 14 16:22 blast/
-rwxrwxrwx 1 onsh onsh 22661379 May 14 16:02 ensGene.txt
-rwxrwxrwx 1 onsh onsh 982498668 May 14 15:12 Homo_sapiens.GRCh38.small.1.bt2
-rwxrwxrwx 1 onsh onsh 842742636 May 14 15:12 Homo_sapiens.GRCh38.small.1.ebwt
-rwxrwxrwx 1 onsh onsh 733719256 May 14 15:12 Homo_sapiens.GRCh38.small.2.bt2
-rwxrwxrwx 1 onsh onsh 366859632 May 14 15:12 Homo_sapiens.GRCh38.small.2.ebwt
-rwxrwxrwx 1 onsh onsh 8945 May 14 15:12 Homo_sapiens.GRCh38.small.3.bt2
-rwxrwxrwx 1 onsh onsh 8945 May 14 15:12 Homo_sapiens.GRCh38.small.3.ebwt
-rwxrwxrwx 1 onsh onsh 733719249 May 14 15:12 Homo_sapiens.GRCh38.small.4.bt2
-rwxrwxrwx 1 onsh onsh 733719249 May 14 15:12 Homo_sapiens.GRCh38.small.4.ebwt
-rwxrwxrwx 1 onsh onsh 3150052344 May 14 15:12 Homo_sapiens.GRCh38.small.fa
-rwxrwxrwx 1 onsh onsh 982498668 May 14 17:15 Homo_sapiens.GRCh38.small.rev.1.bt2
-rwxrwxrwx 1 onsh onsh 842742636 May 14 17:15 Homo_sapiens.GRCh38.small.rev.1.ebwt
-rwxrwxrwx 1 onsh onsh 733719256 May 14 17:15 Homo_sapiens.GRCh38.small.rev.2.bt2
-rwxrwxrwx 1 onsh onsh 366859632 May 14 17:15 Homo_sapiens.GRCh38.small.rev.2.ebwt
-rwxrwxrwx 1 onsh onsh 20767330 May 14 16:02 refGene.txt
-rwxr-xr-x 1 onsh onsh 171175 Jul 23 2015 tophat
-rwxr-xr-x 1 onsh onsh 233 Jul 23 2015 tophat2
-rwxr-xr-x 1 onsh onsh 118420 May 15 13:39 tophat-fusion-post
drwxrwxrwx 0 onsh onsh 512 May 15 15:06 tophat_SRR064286/

In the tophat_SRR064286/ directory, I have some files:

onsh@PC3:~/src/tophat-2.1.0/tophat_SRR064286$ ll
-rw-rw-rw- 1 onsh onsh 2754640914 May 14 15:04 accepted_hits.bam
-rw-rw-rw- 1 onsh onsh 569 May 14 15:04 align_summary.txt
-rw-rw-rw- 1 onsh onsh 1646739 May 14 15:04 deletions.bed
-rw-rw-rw- 1 onsh onsh 16228495 May 14 15:04 fusions.out
-rw-rw-rw- 1 onsh onsh 1380426 May 14 15:04 insertions.bed
-rw-rw-rw- 1 onsh onsh 6025364 May 14 15:04 junctions.bed
drwxrwxrwx 0 onsh onsh 512 May 15 13:58 post_out/ # <— output directory
-rw-rw-rw- 1 onsh onsh 180 May 14 15:04 prep_reads.info
-rw-rw-rw- 1 onsh onsh 75130369 May 14 15:04 unmapped.bam

I dig the tophat-fusion-post code, it seems that “get_chromosome_order()” function doesn’t work.
A sam_header file is empty.

bowtie_header_filename = tmp_dir + "sam_header"
subprocess.call(bowtie_header_cmd,
stdout=open(bowtie_header_filename, "w"),
stderr=open("/dev/null"))

The chrs variable which the function returns is also empty in spite of using chrs variable in the subsequent processes.

Could you anybody solve this?

Thank you all,

onsh

[onsh]

Bowtie indexes are downloaded from Ensembl. In the original file,
I added a word “chr” to each chromosome numbers because the chromosome number is just a number.
onsh

I'm sorry that the sentence above is not correct.

<CORRECTION>
FASTA of genome sequences are downloaded from Ensembl. In the original file, I added a word “chr” to each chromosome numbers because the chromosome number is just number.
Bowtie indexes were built by bowtie 1.2.2.