[stanwish]

Haha, same question here.
I followed the review protocol paper and used almost same command as kenietz, the gene_exp.diff combined several genes together but the isoform_exp.diff is normal.

Does anyone have any idea on it?

Thanks.

Hi guys,
very interesting thread. I have one question though. What about the merging of genes when they are very close on the same strand?

------- from gene_exp.diff ---------
AT1G01225,AT1G01230 Chr1:95986-99240
----------------------------------------

when actually gene AT1G01225 is 95987-97407 and AT1G01230 is 97456-99240. But the merged.gtf says that AT1G01225 is contained in AT1G01230. Cant get that

I have WT and MUT and my flow for both is like that:
>tophat -p 4 -o tophat_wt -G TAIR_9.gff ~/Work_drive/Genomes/TAIR_9/TAIR_9_rg WT-1.fq.gz
>cufflinks -o cuffl_wt -p 4 -g TAIR_9.gff tophat_wt/accepted_hits.bam

Then:
>cuffmerge -p 4 -s TAIR_9_rg.fa -g TAIR_9.gff assemblies.txt
>cuffdiff -p 4 -o cuffdiff_out -L WT,MUT tophat_wt/accepted_hits.bam tophat_mut/accepted_hits.bam

Contents of 'assemblies.txt":
cuffl_wt/transcripts.gtf
cuffl_mut/transcripts.gtf

I saw somewhere on the forum that -g should be used with cufflinks in order to do RABT and solve the issue with merging of genes. But i see its not working so. So now is advised to run cufflink without any -g or -G option and then give the reference annotation to cuffmerge only?

Thank you for you help

[billstevens]

Hi guys,

I have a question on read depth. Cole has mentioned that cuffdiff only tests when there are at least 10 fragments in one of the conditions. Which output file tests that? I was thinking it was genes.count_tracking. However, I checked this when I had only 100M reads versus 200M reads, and both times I ended up with the same number of genes over 10 fragments.

Anyone know how to check read depth?

[swaraj]

I have a question about cuffmerge and cuffcompare. I have three samples and have independently assembled each using cufflinks. I am only interested in intergenic long non-coding RNA transcripts.

My samples are developmental stages. Is there a possibility that if I use cuffmerge to combine the assembled gtf from each sample, I run the risk of merging transcript A expressed in sample1 with transcript B expressed in sample2, where A is contained within B.

I would say I want the two transcripts to be reported separately as probably they have same splicing pattern but alternative promoters during two different stages of development.

[son_nexg]

Hi all,

I ran into a problem while using the cuffmerge. A transcript of my interest was their in three individual cufflinks assemblies (.gtf) but it got lost when I merged them using the cuffmerge. Here is how the transfrags looked post cufflinks step:
9 Cufflinks transcript 3475480 3479343 1000 + . gene_id "A_L004__R1.14999"; transcript_id "Solyc09g010080.2.1"; FPKM "240.6680251912";
9 Cufflinks exon 3475480 3475720 1000 + . gene_id "A_L004__R1.14999"; transcript_id "Solyc09g010080.2.1"; exon_number "1";
9 Cufflinks exon 3477006 3477014 1000 + . gene_id "A_L004__R1.14999"; transcript_id "Solyc09g010080.2.1"; exon_number "2";
9 Cufflinks exon 3477092 3478113 1000 + . gene_id "A_L004__R1.14999"; transcript_id "Solyc09g010080.2.1"; exon_number "3";
9 Cufflinks exon 3478293 3478537 1000 + . gene_id "A_L004__R1.14999"; transcript_id "Solyc09g010080.2.1"; exon_number "4";
9 Cufflinks exon 3478728 3478827 1000 + . gene_id "A_L004__R1.14999"; transcript_id "Solyc09g010080.2.1"; exon_number "5";
9 Cufflinks exon 3478959 3479343 1000 + . gene_id "A_L004__R1.14999"; transcript_id "Solyc09g010080.2.1"; exon_number "6";
9 Cufflinks transcript 3475480 3479343 1000 + . gene_id "B_L004_R1.14955"; transcript_id "Solyc09g010080.2.1"; FPKM "303.9854176646";
9 Cufflinks exon 3475480 3475720 1000 + . gene_id "B_L004_R1.14955"; transcript_id "Solyc09g010080.2.1"; exon_number "1";
9 Cufflinks exon 3477006 3477014 1000 + . gene_id "B_L004_R1.14955"; transcript_id "Solyc09g010080.2.1"; exon_number "2";
9 Cufflinks exon 3477092 3478113 1000 + . gene_id "B_L004_R1.14955"; transcript_id "Solyc09g010080.2.1"; exon_number "3";
9 Cufflinks exon 3478293 3478537 1000 + . gene_id "B_L004_R1.14955"; transcript_id "Solyc09g010080.2.1"; exon_number "4";
9 Cufflinks exon 3478728 3478827 1000 + . gene_id "B_L004_R1.14955"; transcript_id "Solyc09g010080.2.1"; exon_number "5";
9 Cufflinks exon 3478959 3479343 1000 + . gene_id "B_L004_R1.14955"; transcript_id "Solyc09g010080.2.1"; exon_number "6";
9 Cufflinks transcript 3475480 3479343 1000 + . gene_id "C_L004_R1.15064"; transcript_id "Solyc09g010080.2.1"; FPKM "264.2356132737";
9 Cufflinks exon 3475480 3475720 1000 + . gene_id "C_L004_R1.15064"; transcript_id "Solyc09g010080.2.1"; exon_number "1";
9 Cufflinks exon 3477006 3477014 1000 + . gene_id "C_L004_R1.15064"; transcript_id "Solyc09g010080.2.1"; exon_number "2";
9 Cufflinks exon 3477092 3478113 1000 + . gene_id "C_L004_R1.15064"; transcript_id "Solyc09g010080.2.1"; exon_number "3";
9 Cufflinks exon 3478293 3478537 1000 + . gene_id "C_L004_R1.15064"; transcript_id "Solyc09g010080.2.1"; exon_number "4";
9 Cufflinks exon 3478728 3478827 1000 + . gene_id "C_L004_R1.15064"; transcript_id "Solyc09g010080.2.1"; exon_number "5";
9 Cufflinks exon 3478959 3479343 1000 + . gene_id "C_L004_R1.15064"; transcript_id "Solyc09g010080.2.1"; exon_number "6";


but this particular transcript (Solyc09g010080.2) does not exist in the final merged.gtf.

I used -g and -s options with cuffmerge.

Does anyone has a guess why that would happen?

Thanks in advance for your help.
Cheers!

[sdriscoll]

Looking at the dates of the questions, and the lack of responses, in this thread I get the feeling that the cufflinks developers are quietly waiting for people to stop using their tools. This combined with their google user group silence makes me wonder...is there a future for these tools?

[ron128]

Looking at the dates of the questions, and the lack of responses, in this thread I get the feeling that the cufflinks developers are quietly waiting for people to stop using their tools. This combined with their google user group silence makes me wonder...is there a future for these tools?

Amen! Couldnt agree more!

IMHO though, Cufflinks-tophat is still a very viable option, inspite of the developers silence!

[Gonza]

Dear All,

Basic question. I run tuxedo without using cuffcompare. Used tophat2 --> cufflinks --> cuffmerge --> cuffdiff --> R.
Data looks good, makes sense. However, I cannot figure out how to have access to 8.1.1 Gene Feature plots (http://compbio.mit.edu/cummeRbund/manual_2_0.html)

For example, I do :

myGeneId<-"XLOC_012114"
myGene<-getGene(cuff_data,myGeneId)
myGene
XLOC_012114 <-expressionPlot(myGene)
XLOC_012114

head(fpkm(myGene))
head(fpkm(isoforms(myGene))) # up to here all works

head(features(myGene))# not working yet
genetrack<-makeGeneRegionTrack(myGene) # not working yet
plotTracks(genetrack) # not working yet

Is this because i did not use cuffcomare? should I rerun the analysis? Is that critical to use cuffcompare?

Please let me know what you think
Thanks in advance