Tophat-fusion-post outputting no fusions

[spidermonkey]

I'm trying to get fusion post running on human data, yet fusion-post returns 0 fusions. These samples had been run using a different index, with a non-zero amount of fusion results, therefore I'm not sure why this isn't working. The g1k index however needs to be used on these samples, and I'm using the same index for fusion-post as was used for aligning the tophat reads. The directory setup apart from the index should otherwise be identical to the working setup.

My directory structure:

tophat_299aM
tophat_299bN
tophat_299bT

mcl
ensGene.txt
refGene.txt

blast

The blast directory contains symlinks to human_genomic, other_genomic and nt blast files. The index contains bowtie1 built human g1k index files (tophat with fusion was run using bowtie1).

[Tue Jan 15 11:49:19 2013] Beginning TopHat-Fusion post-processing run (v2.0.6)
-----------------------------------------------
[Tue Jan 15 11:49:19 2013] Extracting 23-mer around fusions and mapping them using Bowtie
[Tue Jan 15 11:54:09 2013] Filtering fusions
        Processing: tophat_299aM/fusions.out
        Processing: tophat_299bN/fusions.out
        Processing: tophat_299bT/fusions.out
        0 fusions are output in ./tophatfusion_out/potential_fusion.txt
[Tue Jan 15 11:55:06 2013] Blasting 50-mers around fusions
[Tue Jan 15 11:55:06 2013] Generating read distributions around fusions
[Tue Jan 15 11:55:06 2013] Reporting final fusion candidates in html format
        num of fusions: 0
-----------------------------------------------
[Tue Jan 15 11:55:06 2013] Run complete [00:05:47 elapsed]

I'm using the following parameters to run fusion-post on a cluster with 64GB of allocated memory to the job:

tophat-fusion-post -p 8 --num-fusion-reads 1 --num-fusion-pairs 1 --num-fusion-both 0 bowtie1_g1k_idx/human_g1k_v37

I get this output with 0 potential fusions (the fusions.out files in the sample directories are non-empty):

   
drwxr-xr-x  7 user dir    131072 Jan 15 11:55 .
drwxr-x--- 10 user dir    131072 Jan 16 10:30 ..
drwxr-xr-x  2 user dir    131072 Jan 15 11:55 blast_genomic
drwxr-xr-x  2 user dir    131072 Jan 15 11:55 blast_nt
drwxr-xr-x  2 user dir    131072 Jan 15 11:55 check
-rw-r--r--  1 user dir 984352595 Jan 15 11:53 fusion_seq.bwtout
-rw-r--r--  1 user dir 154987245 Jan 15 11:49 fusion_seq.fa
-rw-r--r--  1 user dir 195266454 Jan 15 11:54 fusion_seq.map
drwxr-xr-x  2 user dir    131072 Jan 15 11:55 logs
-rw-r--r--  1 user dir         0 Jan 15 11:55 potential_fusion.txt
-rw-r--r--  1 user dir   1120412 Jan 15 11:55 result.html
-rw-r--r--  1 user dir         0 Jan 15 11:55 result.txt
-rw-r--r--  1 user dir        18 Jan 15 10:19 sample_list.txt
drwxr-xr-x  2 user dir    131072 Jan 15 11:54 tmp

I feel like I've tried everything suggested in the other threads where people have had the same problem. Any ideas?

[jp.]

Hi
did you get solution ?

[spidermonkey]

Unfortunately, no. Our solution was to use another fusion detector (defuse) instead.

[jp.]

I think I can tell you something cos I have successfully used tophat-fusion.
Please post your commands and directory structure and pwd.