This should fix an issue encountered when working from the cord blood reference RGsets (e.g. FlowSorted.CordBlood.450k, etc.). The former approach retains a single NA in trainingProbes if any were encountered in probeList, which will get passed to p[trainingProbes,] and throw an error. For details, see: https://support.bioconductor.org/p/9148370/
This should fix an issue encountered when working from the cord blood reference RGsets (e.g. FlowSorted.CordBlood.450k, etc.). The former approach retains a single NA in
trainingProbesif any were encountered inprobeList, which will get passed top[trainingProbes,]and throw an error. For details, see: https://support.bioconductor.org/p/9148370/