It hasn't been that long since Simon's post below in this thread: http://seqanswers.com/forums/showthread.php?t=5793 but is there anything new out there in terms of applying microarray time series analysis methods to count data? I'm very interested in collecting this type of time series discrete count data but am new to the statistical methods involved. Could count data be normalized somehow- e.g. by library size- and used directly in EDGE-type models for microarray expression data or is the error model just fundamentally different? I've used DESeq for pairwise comparisons but would like to find something that uses the power of time series data to detect significant differences between digital expression trajectories.
Originally posted by Simon Anders
View Post