I would do them together, though I guess it won't make much of a difference as the quantile chosen will likely fall to default and then you're calculating a sample specific scalar. On 2017-11-17 03:49, lilianeconteville [bioc] wrote: > Activity on a post you are following on support.bioconductor.org [1] > > User lilianeconteville [2] wrote Question: Should I normalize communities separately with metagenomeseq? [3]: > > Hello, > > I have shotgun metagenomics data from fecal samples and my objective is to compare my data with shotgun metagenomic fecal data of three other communities already published. > > For taxonomic analysis, I used Kraken (https://ccb.jhu.edu/software/kraken/) which gives me the raw number of reads that were assigned to each taxon. I am using metagenomeSeq to normalize the Kraken output and be able to compare the samples, however, I am wondering if I should normalize the samples separately, I mean, I wanna compare the gut microbiome of four human communities, so my question is if I should normalize the output of each community separately or not. > > Thank you in advance, > > Liliane > ------------------------- > > Post tags: metagenomeseq, microbiome, gut microbiome, metagenomics, shotgun metagenomics > > You may reply via email or visit Should I normalize communities separately with metagenomeseq? Links: ------ [1] https://support.bioconductor.org [2] lilianeconteville [3] Should I normalize communities separately with metagenomeseq?