cuffcompare output

[dinesh tripathy]

retreiving sequences

i did cuffcompare and the files "cuffcmp.combined.gtf" "cuffcmp.loci" "cuffcmp.stats" "cuffcmp.tracking". i am interested in unknown transcripts. so i did a grep on .tracking file. the count came to 3940.
now, how do i get the sequence of those 3940 transcripts?

[Ajayi Oyeyemi]

and in addition to the above, can someone help me with the cuffcompare output posted above to answer this questions:

Determining the total number of transcripts (Known, partial and novel) assembled that are compatible with the existing annotation.
How to determine the total of unannotated spliced isoforms of known genes.

How to determine the number of transcripts found in the intergenic regions at certain distances like 1,000bp from known genes.

Thanks.

[Ajayi Oyeyemi]

Can someone help me out with this cuffcompare output.
I'm imploring someone to please help or provide a link that will give me more information.

[Ajayi Oyeyemi]

Cuffcompare output

Hi Folks, I need your help on this isssue. I ran cuffcompare and I got the following output below:
#= Summary for dataset: cufflinks_6/transcripts.gtf :
# Query mRNAs : 41274 in 33119 loci (24486 multi-exon transcripts)
# (6071 multi-transcript loci, ~1.2 transcripts per locus)
# Reference mRNAs : 26679 in 24525 loci (20258 multi-exon)
# Corresponding super-loci: 14030
#--------------------| Sn | Sp | fSn | fSp
Base level: 66.3 49.7 - -
Exon level: 48.9 60.9 51.2 63.8
Intron level: 63.4 88.3 63.7 88.8
Intron chain level: 31.1 25.7 39.1 32.3
Transcript level: 0.0 0.0 0.4 0.2
Locus level: 25.4 18.8 29.7 21.9

Matching intron chains: 6303
Matching loci: 6236

Missed exons: 69277/207203 ( 33.4%)
Novel exons: 24838/166289 ( 14.9%)
Missed introns: 59718/181523 ( 32.9%)
Novel introns: 6896/130363 ( 5.3%)
Missed loci: 9351/24525 ( 38.1%)
Novel loci: 14555/33119 ( 43.9%)

Total union super-loci across all input datasets: 35358
(12030 multi-transcript, ~3.8 transcripts per locus)

Can someone help me with the intepretation of this result? I searched through the manual but got no clue.
THanks.

[agent99]

You may also need to change MT to chrM for mito genes.

[Thomas Doktor]

You're right, I was mixing it up with BED format. Apologies.

[sterding]

Hi,

Just to confirm, Ensembl does use 1-based coordinates for the genome (in gtf and other files).

Yes, I agree. I was just correcting Tom's comment of subtracting 1 from Ensembl's GTF. It's not needed.

[Giulietta EnsemblHelpdesk]

Hi,

Just to confirm, Ensembl does use 1-based coordinates for the genome (in gtf and other files).

[sterding]

You also need to convert the chromosome coordinates in the Ensembl GTF from the first base being 1 to first base being 0 (simply subtract 1 from the start coordinate).

hi, Tom

I think Ensembl GTF is fine itself without subtracting 1 from the start coordinate, since GTF start is 1-based.

[Thomas Doktor]

As pointed out, the standard Ensembl GTF file contains chromosome numbers (1,2,3,..) instead of chromosome identifiers (chr1,chr2,chr3,...) so you need to convert these. You also need to convert the chromosome coordinates in the Ensembl GTF from the first base being 1 to first base being 0 (simply subtract 1 from the start coordinate).

After these two conversion steps you should be able to use the Cufflinks suite of programs.

EDIT
You do not need to edit the coordinates in the GTF file.

[Xi Wang]

Hi there

I am running Cufflinks 'cuffcompare' on transcript.gtf (produced by cufflinks) and comparing it with a .gtf file downloaded from Ensemble

as such:

./cuffcompare -r /home/Homo_sapiens.GRCh37.55.gtf transcripts.gtf

and I seem to get no matches at all between files?!

#= Summary for dataset: transcripts.gtf :
# Total mRNAs : 17735 in 17537 loci (17696 multi-exon)
# Reference mRNAs : 99330 in 43502 loci (82822 multi-exon)
# Corresponding super-loci: 0
#--------------------| Sn | Sp | fSn | fSp
Base level: 0.0 0.0 - -
Exon level: 0.0 0.0 0.0 0.0
Intron level: 0.0 0.0 0.0 0.0
Intron chain level: 0.0 0.0 0.0 0.0
Transcript level: 0.0 0.0 0.0 0.0
Locus level: 0.0 0.0 0.0 0.0
Missed exons: 353318/353318 (100.0%)
Wrong exons: 45831/45831 (100.0%)
Missed introns: 272474/272474 (100.0%)
Wrong introns: 28264/28264 (100.0%)
Missed loci: 0/43502 ( 0.0%)
Wrong loci: 0/17537 ( 0.0%)


HAs anyone else tried this - where did you get your reference .gtf from? Ive used this previously when TopHat itself calculated RPKM values, and it worked fine

Thanks

I think the names of chromosomes are not matched.

[nat]

Hi there

I am running Cufflinks 'cuffcompare' on transcript.gtf (produced by cufflinks) and comparing it with a .gtf file downloaded from Ensemble

as such:

./cuffcompare -r /home/Homo_sapiens.GRCh37.55.gtf transcripts.gtf

and I seem to get no matches at all between files?!

#= Summary for dataset: transcripts.gtf :
# Total mRNAs : 17735 in 17537 loci (17696 multi-exon)
# Reference mRNAs : 99330 in 43502 loci (82822 multi-exon)
# Corresponding super-loci: 0
#--------------------| Sn | Sp | fSn | fSp
Base level: 0.0 0.0 - -
Exon level: 0.0 0.0 0.0 0.0
Intron level: 0.0 0.0 0.0 0.0
Intron chain level: 0.0 0.0 0.0 0.0
Transcript level: 0.0 0.0 0.0 0.0
Locus level: 0.0 0.0 0.0 0.0
Missed exons: 353318/353318 (100.0%)
Wrong exons: 45831/45831 (100.0%)
Missed introns: 272474/272474 (100.0%)
Wrong introns: 28264/28264 (100.0%)
Missed loci: 0/43502 ( 0.0%)
Wrong loci: 0/17537 ( 0.0%)


HAs anyone else tried this - where did you get your reference .gtf from? Ive used this previously when TopHat itself calculated RPKM values, and it worked fine

Thanks