Small accepted-hits.bam file and all FPKM values is error

Question

[Ginsea Chen]

Hi ALL

I used tophat and cufflinks to analyze my RNA-seq data, my operation is follows:

$ tophat -p 8 -G genes.gtf -o SRR123456 genome SRR123456_1.fq SRR123456_2.fq
$ cufflinks accepted-hits.bam -G genes.gtf -o result

The problem is FPKM values of all genes in genes.fpkm_tracking file is zero, and then I find accepted-hits.bam is very small about 1M.

I can't find any way to solve it, so I ask for your help.

my tophat.log file:

[2014-07-06 20:57:33] Beginning TopHat run (v2.0.11)
-----------------------------------------------
[2014-07-06 20:57:33] Checking for Bowtie
          Bowtie version:     2.2.3.0
[2014-07-06 20:57:33] Checking for Samtools
        Samtools version:     0.1.19.0
[2014-07-06 20:57:33] Checking for Bowtie index files (genome)..
[2014-07-06 20:57:33] Checking for reference FASTA file
[2014-07-06 20:57:33] Generating SAM header for genome
[2014-07-06 20:57:44] Reading known junctions from GTF file
[2014-07-06 20:57:47] Preparing reads
     left reads: min. length=76, max. length=76, 27091797 kept reads (1057227 discarded)
[2014-07-06 21:03:27] Building transcriptome data files SRR392837/tmp/genes
[2014-07-06 21:03:49] Building Bowtie index from genes.fa
[2014-07-06 21:15:00] Mapping left_kept_reads to transcriptome genes with Bowtie2 
[2014-07-06 21:51:47] Resuming TopHat pipeline with unmapped reads
[2014-07-06 21:51:47] Mapping left_kept_reads.m2g_um to genome genome with Bowtie2 
[2014-07-07 00:03:34] Mapping left_kept_reads.m2g_um_seg1 to genome genome with Bowtie2 (1/3)
[2014-07-07 00:30:20] Mapping left_kept_reads.m2g_um_seg2 to genome genome with Bowtie2 (2/3)
[2014-07-07 00:35:32] Mapping left_kept_reads.m2g_um_seg3 to genome genome with Bowtie2 (3/3)
[2014-07-07 02:10:52] Searching for junctions via segment mapping
[2014-07-07 02:28:58] Retrieving sequences for splices
[2014-07-07 02:29:33] Indexing splices
[2014-07-07 02:29:54] Mapping left_kept_reads.m2g_um_seg1 to genome segment_juncs with Bowtie2 (1/3)
[2014-07-07 02:37:24] Mapping left_kept_reads.m2g_um_seg2 to genome segment_juncs with Bowtie2 (2/3)
[2014-07-07 02:41:00] Mapping left_kept_reads.m2g_um_seg3 to genome segment_juncs with Bowtie2 (3/3)
[2014-07-07 03:29:21] Joining segment hits
[2014-07-07 04:07:49] Reporting output tracks
-----------------------------------------------
[2014-07-07 04:13:56] A summary of the alignment counts can be found in SRR392837/align_summary.txt
[2014-07-07 04:13:56] Run complete: 07:16:22 elapsed

[Ashutosh Pandey]

Can you paste the content of the file: SRR392837/align_summary.txt

[2014-07-07 04:13:56] A summary of the alignment counts can be found in SRR392837/align_summary.txt

[Ginsea Chen]

Hi Pandey

My align_summary.txt as follows:

Reads:
          Input     :  28149024
           Mapped   :      1868 ( 0.0% of input)
            of these:      1546 (82.8%) have multiple alignments (436 have >20)
 0.0% overall read mapping rate.

[Ashutosh Pandey]

See that's the problem. You have almost no reads getting aligned to the reference genome. As a result, you see zero FPKM values for the genes. The low mapping rate can be due to

1. using a wrong reference genome to align your data
2. The library type for your data doesn't match with the default Tophat library type or the one you have selected.

Normally, 70-80% of input reads get mapped to the reference genome in case of mammalian genome.

[Ginsea Chen]

Hi

I am sure that my RNA-seq data is correspondent with reference genome, so the error may be due to the second reason. Thank you ~