Tophat-Fusion Error 'Long_Spanning_Reads'

Question

[samsara]

I have run Tophat fusion with paired-end RNA-Seq data, but my run was unsuccessful. The error says running 'long_spanning_reads'. I found a post in SEQanswers, but it did not help much.

Following is the executed command.

tophat -o $DIRECTORY$SAMPLENAME \
-p 16 \
--fusion-search \
--keep-fasta-order \
--bowtie1 \
--no-coverage-search \
--mate-inner-dist 130 \
--library-type fr-unstranded \
$BOWTIEINDEX $READ1 $READ2

Following is the log file from Tophat run.

[2013-02-14 16:05:01] Beginning TopHat run (v2.0.6)
-----------------------------------------------
[2013-02-14 16:05:01] Checking for Bowtie
          Bowtie version:     0.12.7.0
[2013-02-14 16:05:01] Checking for Samtools
        Samtools version:     0.1.12.a
[2013-02-14 16:05:01] Checking for Bowtie index files
[2013-02-14 16:05:01] Checking for reference FASTA file
[2013-02-14 16:05:01] Generating SAM header for /app/references/iGenomes/homo_sapiens/hg19/UCSC/hg19/Sequence/BowtieIndex/genome
    format:         fastq
    quality scale:     phred33 (default)
[2013-02-14 16:05:24] Preparing reads
     left reads: min. length=101, max. length=101, 219973253 kept reads (259655 discarded)
    right reads: min. length=101, max. length=101, 219624658 kept reads (608250 discarded)
[2013-02-14 18:46:25] Mapping left_kept_reads to genome genome with Bowtie 
[2013-02-14 22:37:21] Mapping left_kept_reads_seg1 to genome genome with Bowtie (1/4)
[2013-02-14 23:39:33] Mapping left_kept_reads_seg2 to genome genome with Bowtie (2/4)
[2013-02-15 00:38:36] Mapping left_kept_reads_seg3 to genome genome with Bowtie (3/4)
[2013-02-15 01:41:39] Mapping left_kept_reads_seg4 to genome genome with Bowtie (4/4)
[2013-02-15 02:35:54] Mapping right_kept_reads to genome genome with Bowtie 
[2013-02-15 06:35:33] Mapping right_kept_reads_seg1 to genome genome with Bowtie (1/4)
[2013-02-15 07:32:53] Mapping right_kept_reads_seg2 to genome genome with Bowtie (2/4)
[2013-02-15 08:37:51] Mapping right_kept_reads_seg3 to genome genome with Bowtie (3/4)
[2013-02-15 09:43:46] Mapping right_kept_reads_seg4 to genome genome with Bowtie (4/4)
[2013-02-15 10:39:15] Searching for junctions via segment mapping
[2013-02-15 12:55:31] Retrieving sequences for splices
[2013-02-15 12:58:18] Indexing splices
[2013-02-15 13:19:49] Mapping left_kept_reads_seg1 to genome segment_juncs with Bowtie (1/4)
[2013-02-15 15:32:19] Mapping left_kept_reads_seg2 to genome segment_juncs with Bowtie (2/4)
[2013-02-15 17:49:12] Mapping left_kept_reads_seg3 to genome segment_juncs with Bowtie (3/4)
[2013-02-15 20:05:54] Mapping left_kept_reads_seg4 to genome segment_juncs with Bowtie (4/4)
[2013-02-15 22:14:28] Joining segment hits
    [FAILED]
Error running 'long_spanning_reads':Loading fusions...done

What could be the reason with this error ? I will greatly appreciate your help. Thanks !!

[Nicolas Rosewick]

What is your configuration (RAM) ?

[samsara]

I have 12GB of RAM.

[Nicolas Rosewick]

I think the problem is there. Did you check the state of your process with top ? tophat with fusion-search and 200M reads used more than 12Gb I guess. Check the memory with top.

[samsara]

Thanks a lot. It was truly memory issue. I splitted FASTQ files into two and executed tophat fusion. It worked without reporting any error. Please post above comment as answer :D

Answer 1

[Nicolas Rosewick]

I think the problem is there. Did you check the state of your process with top ? tophat with fusion-search and 200M reads used more than 12Gb I guess. Check the memory with top.