I am working on time course data with replicates. My dataset looks like the following: WT: 6 timepoints with 3 replicates each KO: 6 timepoints with 3 replicates each RE: : 6 timepoints with 3 replicates each Total: 54 samples. I have the following questions:

1. While performing WGCNA, should I do it for separate genotypes or should do it together. Note: The data available in under this dataset is TMM normalised and given in three separate tsv files.
2. Can I use this TMM normalised data for outlier detection or should I directly go for network construction (i suppose there is no normalization step required, as it is already normalized).
3. The gene ids had some duplicates, i removed the duplicates by taking the entries with max expression values, is this the right way to do it.
4. Should I perform any other steps before constructing the network.
5. When I performed WGCNA directly with the TMM normalized data, I ended up getting 49 modules (soft_threshold=18, mergecutheight=0.25), is this normal? It would be really helpful if I someone answers these questions. Please provide some good resources or tutorials for the same. Thank you

This support site is meant to help people with technical questions about using Bioconductor packages, rather than as a place to get analysis advice. WGCNA isn't a Bioconductor package, and you are asking for analysis advice, so your questions are quite off-topic for the site. You might consider asking the questions over on biostars.org instead.