cuffcompare - wrong results

[adrian]

Hi:
I did following:

cuffcompare -i gtf_out_list -r /UCSC/hg19/Annotation/Genes/genes.gtf


cut -f4 cuffcmp.tracking | sort | uniq -c
40816 =
20 c

As per cuffcompare manual column 5 is listed as code column, however I see codes in column 4.

Also my other MAIN concern is why I have over 40K '=' and very little C and no other codes.

Is something wrong with my cufflinks, cuffmerge and cuffdiff?

I also checked same with individual .tmap files:
cut -f3 cuffcmp.transcripts.gtf.tmap | sort | uniq -c
40868 =
20 c
1 class_code


my refmap file is empty.


Could any one suggest what could be wrong. I have been struggling with this for last 4 months and I think I cant get past of this. I request your help.

Thanks
Adrian




The following is what I did :

### TOPHAT ###
tophat2 --b2-fast -p 16 -r 200
-G genes.gtf
UCSC/hg19/Sequence/Bowtie2Index/genome/
-o out
r1.fastq r2.fastq


#### CUFFLINKS #####

cufflinks -p 8 -N -o CLRes
-G genes.gtf
-b UCSC/hg19/Sequence/Bowtie2Index/genome.fa
-M rrna.gtf (this includes rrna and trna)
-u
accepted_hits.bam


####CUFFMERGE ####

cuffmerge -p 16 -o CMergeOut
-g genes.gtf
-s UCSC/hg19/Sequence/Bowtie2Index/genome.fa
runs_list

cuffdiff:
cuffdiff -p 12 -o cuffdiffOut

-N
-u

-b UCSC/hg19/Sequence/Bowtie2Index/genome.fa
-M rrna.gtf (this includes rrna and trna)

-L samp1,samp2,samp3

s11.bam,s12.bam,s13.bam s21.bam,s22.bam,s23.bam s31.bam,s32.bam,s33.bam

[adrian]

I forgot to mention,
I have many differentially expressed genes and spliceforms, I am not sure to believe them or not!
Adrian

[adrian]

sorry for bumping this.
Tophat/cufflinks experts - appreciate your opinion, ideas/suggestions.
thanks
adrian