cuffdiff for groups

[bvk]

I want run cuffdiff for two groups. each has 3 samples. Is the below code right?

cuffdiff -o diff_out -b ../genome/ce10.fa -p 2 -L em,la -u merged_asm/merged.gtf cond1.bam1,cond1.bam2,cond1.bam3 cond1.bam1,cond1.bam2,cond1.bam3

[GenoMax]

No. That is not correct.

-L has to be a comma separated list of condition labels.

-L cond1,cond2 (e.g. -L Control,Mutant)

Your list of bam files (comma delimited names no spaces, then two conditions separated by a space) will be. e.g. cond1.bam1,cond1.bam2,cond1.bam3 (space here) cond2.bam1,cond2.bam2,cond2.bam3

[bvk]

No. That is not correct.

-L has to be a comma separated list of condition labels.

-L cond1,cond2 (e.g. -L Control,Mutant)

Your list of bam files (comma delimited names no spaces, then two conditions separated by a space) will be. e.g. cond1.bam1,cond1.bam2,cond1.bam3 (space here) cond2.bam1,cond2.bam2,cond2.bam3

You missed it. please, see it after 3rd bam file I haven't given any comma. And it is separated with space too.

[GenoMax]

The -L option you have used in your command is not two group names for three samples in each group.

", ../tophat/la/SRR493365_thout/accepted_hits.bam " It also looks like there is a space between , and the .. but that may be an illusion in the way the browser is displaying it.

[dpryan]

Aside from your syntax problems, you have a more serious issue in that you're not specifying the actual experiment correctly. SRR493359, SRR493360 and SRR493361 are from the same sample and should just be merged together into a single BAM file. Similarly SRR493363, SRR493364 and SRR493365 are from the same sample. So, you actually have a 1 vs. 1 sample comparison. Do NOT lump each of the files for each sample into a group, since then you're making fake replicates and will have largely meaningless results (of course, you're doing a 1 vs. 1 comparison, so the results aren't exactly robust to begin with).

[bvk]

The -L option you have used in your command is not two group names for three samples in each group.

", ../tophat/la/SRR493365_thout/accepted_hits.bam " It also looks like there is a space between , and the .. but that may be an illusion in the way the browser is displaying it.

ok. now the below code looks fine I guess

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L SRR493359,SRR493360,SRR493361,SRR493363,SRR493364,SRR493365 -u merged_asm/merged.gtf ../tophat/em/SRR493359_thout/accepted_hits.bam,../tophat/em/SRR493360_thout/accepted_hits.bam,../tophat/em/SRR493361_thout/accepted_hits.bam ../tophat/la/SRR493363_thout/accepted_hits.bam,../tophat/la/SRR493364_thout/accepted_hits.bam,../tophat/la/SRR493365_thout/accepted_hits.bam

[dpryan]

ok. now the below code looks fine I guess

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L SRR493359,SRR493360,SRR493361,SRR493363,SRR493364,SRR493365 -u merged_asm/merged.gtf ../tophat/em/SRR493359_thout/accepted_hits.bam,../tophat/em/SRR493360_thout/accepted_hits.bam,../tophat/em/SRR493361_thout/accepted_hits.bam ../tophat/la/SRR493363_thout/accepted_hits.bam,../tophat/la/SRR493364_thout/accepted_hits.bam,../tophat/la/SRR493365_thout/accepted_hits.bam

No, it's not. You're making two groups with the BAM files and are giving those two groups 6 labels rather than 2. "-L larval,early" would make more sense, though see my earlier reply.

[bvk]

Aside from your syntax problems, you have a more serious issue in that you're not specifying the actual experiment correctly. SRR493359, SRR493360 and SRR493361 are from the same sample and should just be merged together into a single BAM file. Similarly SRR493363, SRR493364 and SRR493365 are from the same sample. So, you actually have a 1 vs. 1 sample comparison. Do NOT lump each of the files for each sample into a group, since then you're making fake replicates and will have largely meaningless results (of course, you're doing a 1 vs. 1 comparison, so the results aren't exactly robust to begin with).

so you say that it should look like this:

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L SRR493359,SRR493360,SRR493361,SRR493363,SRR493364,SRR493365 -u merged_asm/merged.gtf ../tophat/em/SRR493359_60_61_thout/accepted_hits.bam ../tophat/em/SRR493363_64_65_thout/accepted_hits.bam

SRR493359_60_61_thout which has merged bam file of 59,60 and 61
SRR493363_64_65_thout which has merged bam file of 63,64 and 65

[GenoMax]

so you say that it should look like this:

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L SRR493359,SRR493360,SRR493361,SRR493363,SRR493364,SRR493365 -u merged_asm/merged.gtf ../tophat/em/SRR493359_60_61_thout/accepted_hits.bam ../tophat/em/SRR493363_64_65_thout/accepted_hits.bam

SRR493359_60_61_thout which has merged bam file of 59,60 and 61
SRR493363_64_65_thout which has merged bam file of 63,64 and 65

Arrange the labels (larval/early) according to their correspondence with the BAM files.

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L larval,early -u merged_asm/merged.gtf ../tophat/em/SRR493359_60_61_thout/accepted_hits.bam ../tophat/em/SRR493363_64_65_thout/accepted_hits.bam

[bvk]

Arrange the labels (larval/early) according to their correspondence with the BAM files.

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L larval,early -u merged_asm/merged.gtf ../tophat/em/SRR493359_60_61_thout/accepted_hits.bam ../tophat/em/SRR493363_64_65_thout/accepted_hits.bam

Yes, I understood. Thank you very much !!

[bvk]

No, it's not. You're making two groups with the BAM files and are giving those two groups 6 labels rather than 2. "-L larval,early" would make more sense, though see my earlier reply.

Thankyou !! I got it.

[bvk]

No, it's not. You're making two groups with the BAM files and are giving those two groups 6 labels rather than 2. "-L larval,early" would make more sense, though see my earlier reply.

As you said If I run the following command

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L larval,early -u merged_asm/merged.gtf ../tophat/em/SRR493359_60_61_thout/accepted_hits.bam ../tophat/em/SRR493363_64_65_thout/accepted_hits.bam

Now, If I want to find the expression levels for each sample I guess it is not possible, cz of the merged bam file.

Is it possible to give labels as larval,early and giving bam files of 6 samples. Will this work?

for eg:

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L embryo,larva -u merged_asm/merged.gtf ../tophat/em/SRR493359_thout/accepted_hits.bam,../tophat/em/SRR493360_thout/accepted_hits.bam,../tophat/em/SRR493361_thout/accepted_hits.bam ../tophat/la/SRR493363_thout/accepted_hits.bam,../tophat/la/SRR493364_thout/accepted_hits.bam,../tophat/la/SRR493365_thout/accepted_hits.bam

[dpryan]

You only have 2 samples anyway. Any sort of metric you'd get from each of the subfiles isn't terribly meaningful unless you're interested in looking at technical variance.

[bvk]

Arrange the labels (larval/early) according to their correspondence with the BAM files.

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L larval,early -u merged_asm/merged.gtf ../tophat/em/SRR493359_60_61_thout/accepted_hits.bam ../tophat/em/SRR493363_64_65_thout/accepted_hits.bam

As you said If I run the following command

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L larval,early -u merged_asm/merged.gtf ../tophat/em/SRR493359_60_61_thout/accepted_hits.bam ../tophat/em/SRR493363_64_65_thout/accepted_hits.bam

Now, If I want to find the expression levels for each sample I guess it is not possible, cz of the merged bam file.

Is it possible to give labels as larval,early and giving bam files of 6 samples. Will this work?

for eg:

cuffdiff -o diff_out4 -b ../genome/ce10.fa -p 2 -L embryo,larva -u merged_asm/merged.gtf ../tophat/em/SRR493359_thout/accepted_hits.bam,../tophat/em/SRR493360_thout/accepted_hits.bam,../tophat/em/SRR493361_thout/accepted_hits.bam ../tophat/la/SRR493363_thout/accepted_hits.bam,../tophat/la/SRR493364_thout/accepted_hits.bam,../tophat/la/SRR493365_thout/accepted_hits.bam

[GenoMax]

Technically cuffdiff command as outlined above will work. But Devon has already warned you about the consequence in post #5.