I seek help from forum members in identifying the most appropriate analysis tools / software for
performing the following analyses on the following time-series NGS RNA-Seq count data (total of 144 libraries):

  • 9 time points,
  • 4 genotypes, and
  • 4 replicates each.

The following BioStars posts - A, B, C , D, E did help a little bit, but several of them are sort of old,
and I wonder if there are better, up-to-date tools that are more widely used these days.

I've filled in answers to some of my questions, but please do add / comment if you think any of those
answers should be different.

Q1. Differential Gene Expression (DGE) across entire time series, between genotypes

Could it be just DESeq2, based on these BioStars posts - older
, newer, and on bioconductor

Q2. DGE at a specific time point, between genotypes

Just DESeq2 (pairwise) ?

Q3. DGE between different time points, within same genotype

Again, just DESeq2 (pairwise comparison) ?

Q4. Reporting Gene Co-Expression Networks (GCNs) for time-series data from one genotype (let's call this baseline)

WGCNA, TimeDelay-ARACNE, PropaNet ?

Q5. Reporting which GCNs are broken / altered in 'test' genotype vs. 'baseline' genotype

Q6. Reporting temporal shifts in expression of GCN components in test genotype vs. baseline data

Q7. Reporting quantitative shifts in expression of GCN components in test genotype vs. baseline data

Q8. Some genes are going to be cyclic and they could confound inferences, right? How is this controlled?

Q9. And some other genes will be shifted in expression strength and/or timing, and these could also confound analyses? Again, how is this problem solved?

Open ended question: Furthermore, what are other considerations to keep in mind while performing time-series analyses and for choosing the correct tools for answering questions 1 - 7 listed above?

I am starting to practice using DESeq2 and WGCNA on my read counts, but I am not entirely sure these are the
best choices to answer my questions. I look forward to all your replies. Thanks again!



Source link