Batch effect correction without QC sample

[Kotchani]

I've encountered an issue with my samples that were run in different batches. I've been following the instructions provided on this website: https://bioconductor.org/packages/devel/bioc/vignettes/pmp/inst/doc/pmp_vignette_sbc_spectral_quality_assessment.html. However, I've run into a challenge because I don't have a QC sample. I attempted to use samples from batch 1 as a QC substitute, but during the for-loop step to store RSD, I noticed that the values in the matrix for some columns are all NA.

RSD_QC (batch 1)

RSD_sample

What would you recommend I do in this situation? Alternatively, if I don't have a QC sample, is there another approach to do batch correction?

Your guidance would be greatly appreciated.

[sneumann]

Hi, indeed that's not an xcms issue. Please check out e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4796354/ which has an overview, including "strategies based on the study samples". Yours, Steffen