Question: Identify Differentially Expressed Genes With Two Groups In Which There Are Only Two Replicates
1
gravatar for liux.bio
6.4 years ago by
liux.bio340
China
liux.bio340 wrote:

Hello everyone,

I want to Identify differentially expressed genes with two groups in which there are only two replicates.

What methods should I use? I want to use t-statistics or Mann- Whitney test, but I am not sure the minimum replicates of the t-statistics or Mann- Whitney test.

Thank you

ADD COMMENTlink modified 6.4 years ago by Steve Lianoglou5.0k • written 6.4 years ago by liux.bio340
3

Can you clarify your question? You say you have two replicates, which suggests that you have four samples (2 groups x 2 replicates = 4). Or do you have two groups with 0 replicates -- ie. you have 2 samples.

Also, your tag says "microarray" but it is not mentioned in your question what platform you have used for expression measurement. Can you please further modify your question to specify if that this is microarray (vs. rna-seq, for example)?

ADD REPLYlink written 6.4 years ago by Steve Lianoglou5.0k

Ok.I have four samples(2 groups x 2 replicates).The platform I used is Affymetrix GeneChip Mouse Genome 430A 2.0. Could I use Mann- Whitney test?

ADD REPLYlink written 6.4 years ago by liux.bio340
2

well minimum sample size for t-test is 2 (for which the degree of freedom = 1), but the statistics won't be very reliable.

ADD REPLYlink written 6.4 years ago by Bharat Iyengar270

Thank you! What about Mann- Whitney test?

ADD REPLYlink written 6.4 years ago by liux.bio340

it can be performed too. but in general smaller the sample size, less reliable will be the result. .

ADD REPLYlink written 6.4 years ago by Bharat Iyengar270
1

What kind of replicates? Biological replicates? Technical replicates?

If you only have technical replicates, then it does not make sense to use a statistical test. What you will pick up will be technical differences, and not biological differences. Even if they are biological replicates, I would be very skeptical of ANY results produced with only 2 replicates. It may work well for a pilot study to asses whethere there was enough RNA in a given sample or likewise, but the biological conclusions you can draw from such a dataset if very limited.

ADD REPLYlink written 6.4 years ago by David Westergaard1.4k

Many thanks for your advice.

ADD REPLYlink written 6.4 years ago by liux.bio340
1
gravatar for Steve Lianoglou
6.4 years ago by
Steve Lianoglou5.0k
US
Steve Lianoglou5.0k wrote:

Unless you have a very good reason not to do so, your first preference for any microarray analysis should always be limma.

There is an extensive user's guide which is very much worth reading, as well, before you start your analysis.

ADD COMMENTlink written 6.4 years ago by Steve Lianoglou5.0k

Thank you.I have read it.

ADD REPLYlink written 6.4 years ago by liux.bio340
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1967 users visited in the last hour