I have a RNA-Seq time-series data (vertebrate development with 9 stages and 2 samples per stage). So far I mapped these reads to the genome using BWA and used these alignments to count the number of reads on the level of genes using HTSeq count since I'm not interested in exons but complete genes. I'm using these counts to first calculate frequencies of expression value for each gene, (e.g., count of particular gene/sum of all counts in a sample), and then to calculate sum of frequencies for each sample and stage. As a result of this procedure I get the cumulative values (cumulative frequency) that show how the total expression varies between the stages. The question: I want to try different normalization techniques (on the HTSeq count results) to see how they affect my cumulative expression values. How to do only the normalization(s) without the differential expression of genes that usually comes later in packages?. What are my options among the packages available?