Any suggestion of tool for differential expression test for my situation?

[thejustpark]

I am trying to get differentially expressed genes for this experiment*
where one condition has one biological replicate (two samples) and the other condition doesn't have one (one sample).*
I read many discussions about how bad it is to do differential expression test for no replicate samples.*

But I am still unclear how to do the test for my situation (I believe my situation is better than completely no replicate case).*
Can you guys suggest me a good tool for this analysis?

Thanks,*
HJ

[thejustpark]

Just one more comment: the biologists said that the two conditions are expected to be not much different, only affecting a number of genes. That is another reason I think it would be OK not to require additional replicate at least at this time point.

[dariober]

I think there is no "right" way of addressing this situation. I would use edgeR or DEseq as they implement methods designed to make the best of few replicates, although your case (2 + 1) is quite extreme.

Rather you should interpret the results in the light of such small sample sizes. If you have some genes that are expected to change, see how they behave and see if the genes detected as different make sense in biological terms.

That is another reason I think it would be OK not to require additional replicate at least at this time point.

Mmm... I'd say the opposite. If the effect size you want to detect is small you should put more replicates.

[dpryan]

@thejustpark: Please don't cross-post on here and biostars, it just doubles the work since you get the same answer from both places.

[thejustpark]

Thank you Devon for pointing that out.
I didn't know. I will keep that in my mind.

HJ.

[Xuegong]

I am trying to get differentially expressed genes for this experiment*
where one condition has one biological replicate (two samples) and the other condition doesn't have one (one sample).*
I read many discussions about how bad it is to do differential expression test for no replicate samples.*

But I am still unclear how to do the test for my situation (I believe my situation is better than completely no replicate case).*
Can you guys suggest me a good tool for this analysis?

Thanks,*
HJ

How about doing two comparisons of each of the replicate with no-replicate sample and then check the common findings from the two comparisons? This may help to make the results more reliable. For comparing two single samples, you may check out the DEGseq tool. It's simple to use and can give reasonable results, especially when you don't have replicates to estimate more complex models. Also you may do a replicate-vs-replicate comparison first as the negative control of the difference distribution, and use this distribution to judge the significance of your real comparisons. In this way, you avoid the use of any model and parameter estimation.