That error means you have samples without a single positive count. Use the function filterData to filter them out before normalization. On Aug 6, 2016, 1:17 PM, at 1:17 PM, "dr.richrodrigues [bioc]" <noreply@bioconductor.org> wrote: >dr.richrodrigues posted the Question: "Metagenome seq errors using biom >(1.3) file": > >Hi I generated a biom (1.3) file using QIIME (1.8). I am trying to run >metagenomeseq in R (biomformat_1.0.2 metagenomeSeq_1.14.2). I can read >the biom file but get errors in the following steps. Help appreciated! >> x = read_biom("miseq10/otu_table.fractn0.001.biom") > ObjData = >biom2MRexperiment(x) > ObjData MRexperiment (storageMode: >environment) assayData: 79 features, 160 samples element names: counts >protocolData: none phenoData: none featureData featureNames: 184451 >3957916 ... 259556 (79 total) fvarLabels: taxonomy1 taxonomy2 ... >taxonomy7 (7 total) fvarMetadata: labelDescription experimentData: use >'experimentData(object)' Annotation: > obj = cumNorm(ObjData, p = >cumNormStatFast(ObjData)) Error in `[<-`(`*tmp*`, leng:(leng - >length(smat[[i]]) + 1), i, value = numeric(0)) : subscript out of >bounds > obj = cumNorm(ObjData, p = cumNormStat(ObjData)) Error in >cumNormStat(ObjData) : Warning empty sample > >--- >See the full post at: Metagenome seq errors using biom (1.3) file >Replying to this email will post an answer to the question above.