Statistical test for unbalanced data set

[owomo]

Hi, guys,
My interest is to ask whether a gene is differentially expressed between group A and group B.
but the experiment design is unbalanced as follow:

A1, A2, A3, B1_repeat1, B1_repeat2, B2_repeat1, B2_repeat2.

So, group A contains 3 samples, but group B contains only two samples, and each have two biological replicates.

I have used the edgeR to fit a generalized linear models (GLMs). But a review says that it may be not appropriate to manage the biological replicates in group B in this way.
My questions are:
What is the most proper way to manage the biological replicates? Is a mixed effect model suitable is this case?
Thank your very much for you time, any suggestion would be greatly appreciated.

[dpryan]

edgeR works fine for unbalanced designs. One question is what the repeats in group B represent. Are these re-runs of the same library (i.e., they represent a minimal batch effect) or were they made from a different library. Also, did the first try at samples B1 and B2 just produce crappy results or why else did you repeat them?

[owomo]

edgeR works fine for unbalanced designs. One question is what the repeats in group B represent. Are these re-runs of the same library (i.e., they represent a minimal batch effect) or were they made from a different library. Also, did the first try at samples B1 and B2 just produce crappy results or why else did you repeat them?

Thank you for your reply. Sorry I didn't clarify my experiment.
The replicates in group B is biological replicates, it is not re-runs of the same
for example,
In group 1, we have species A, species B, species C.
In group 2, we have species D, with individual D1 and D2, and species E, with individual E1 and E2.
We would like to test whether a gene is significantly differentially expressed between group1 (including species A, B, C) and group2 (including species D, E).
So any suggestions? Thank you very much.

[dpryan]

Right, I'm asking about B1_repeat1 vs. B1_repeat2. Are those actually biological replicates or are they technical replicates. If the latter, why did you bother producing them?

[owomo]

Right, I'm asking about B1_repeat1 vs. B1_repeat2. Are those actually biological replicates or are they technical replicates. If the latter, why did you bother producing them?

Sorry, I edit my last thread again and hope it is clear now.

In group 1, we have species A, species B, species C.
In group 2, we have species D, with individual D1 and D2, and species E, with individual E1 and E2.

So there are 2 sample in group 2, each with 2 biological replicates, while in group 1, we have 3 sample, but no replicates.

1, dividual D1 and D2, individual E1 and E2 are biological replicates, not technical replicates.
2, I wanna to test group 1 (including species A,B,C) Vs. group 2 (including species D, E).

Many thanks for your time

[owomo]

OK, I think I can simplify this model as follows,
group 1 have 3 biological replicates,
group 2 have 2 biological replicates, but each with 2 technical replicates.
for example:

group exp
A1 23
A2 25
A3 27
B1.1 89
B1.2 85
B2.1 78
B2.2 82

So, how to manage the technical replicates in group 2?

Thanks a lot