To the developers,

I have an RNA-Seq data from a reproductive-stage drought stress in rice. I have 2 genotypes (tolerant and sensitive) and 2 conditions (drought and well-watered) with 4 replications each. Basically, a 2x2 factorial experiment with 4 replications thus having 16 samples. I am working on 2 different tissues (flag leaf and emerging panicles). The analysis was done on each tissue sample.

I used DESEq2 to normalize the expression values after pre-filtering them as suggested in DESEq2. I ahve about 18359 genes for flag leaf and 17561 for emerging panicles. 

I did WGCNA for each tissue using signed hybrid as the network type and pearson as the corType with maxBlocksize of 20000 and minBlocksize of 20. For the mergedCutHeight i tried using 0.2 and 0.1 for each tissue sample.

You mentioned in one of the threads that "in practice, on data sets with 50-100 samples, using 0.15 to 0.2 has worked fairly well. For fewer samples a larger valuse (0.25 to 0.3) may be warranted." My question then is, is it better to use the 0.2 value as the better mergedCutHeight value than 0.1 considering the sample size of my dataset?

Thanks and best regards

Asher