Cuffdiff Samples and replicate arrangement

[Zapages]

I have simple question concerning replicates and samples.

I have 4 experiment sets that comparing two organs.

Experiment A has:
Organ X - 2 replicates
Organ Y - 2 replicates

Experiment B has
Organ X - 2 replicates
Organ Y - 3 replicates.

Currently in Cuffdiff with False Discovery Rate 0.1, I have set the following sample names:

Experiment A - Organ X with 2 Replicates
Experiment A - Organ Y with 2 Replicates
Experiment B - Organ X with 2 Replicates
Experiment B - Organ Y with 3 Replicates

Is it fine if I do the following:

Organ X - 4 Replicates
Organ Y - 5 Replicates

Is it fine to combine the replicates across the two experiments together and just compare the 2 organs. Currently I am getting good P and Q values.


Thank you in advance.

-Zapages

[TiborNagy]

I suggest not to combine the replicates, because the differential expression analysis can use stronger statistics if you have separate replicates.

[Zapages]

I suggest not to combine the replicates, because the differential expression analysis can use stronger statistics if you have separate replicates.

Thank you, I have follow on question. What if I have a dataset like this:

Experiment A:
Organ X - 1 replicate
Organ Y - 2 replicates

Experiment B:
Organ X - 1 replicate

Is it fine if I combine the replicates here for cuffdiff or leave them separate? will that even work?

Sorry for these questions. Thank you again.