Label for every BAM in Cuffdiff

[sindrle]

Hi!
Im running Cuffdiff 2.1.1 now on some samples, but Im struggling with the --labels option.

It allows me only to name the conditions, but I really want to name the individual BAMs.

I want to do this because its easier when doing downstream analyses.


Anyone knows? Thanks.

[sindrle]

Or, if not possible to label individual BAMs, how do I know which BAMs corresponds to cummerbunds "sample_names"?

[sindrle]

The most tedious way must to manually rename using "find & replace" in all the "read_group_tracking" - files...


But still, the replicates have an odd organisation, so I don't know which BAM correspond to which replicate.

replicate
1
0
2
3
4
6
5
7
8
18
16
9
10
11
12
13
14
15
17
19
20
21
22
23

[sindrle]

I guess nobody knows?

I have emailed the Cuffdiff authors, if they answer Ill post it here.

[TiborNagy]

If you would like to use the bam file names instead of conditions, you loose some statistical power, bacuse you can not handle replicates. If you do not have replicates, you can use the bam file name as label.

[sindrle]

Well, I don't know if I understand you correctly.

Of course Ill use all nye replicates, its just I want to know which BAM has which name in the read-group-files.

If I have 5 replicates in each condition, I want to know that BAM 1 = condition 1 & replicate name 1. etc.

In Cuffdiffs output files this is however impossible to know. Condition is ok, but the replicate name is very confusing.

[sindrle]

Seriously? Nobody has wondered the same??

[sindrle]

Got it!

In Cuffdiff 2.1.1 it gives a "read_groups.info" file. Finally...