Hi ! I saw that what I want to do (multiple tumor samples per patient) should be in the next versions of oncoanalyser (thanks !), but in the meantime we have to find other solutions. I'm running an analysis on patients with multiple tumor samples (diagnosis and relapse). I ran it the first time by using one group_id per tumor sample, but it duplicates the alignment for the normal samples. You recommended to first run the alignement and redux for each sample, then run the analysis from the bams. But I'm stuck because

germline analysis only is not supported

, even if I'm only running alignement and redux. I could pair the normal sample with one tumor sample for the alignment and it would work (just a bit tideous). Is there a way to bypass this verification and perform the alignement on a normal sample alone ? How do you design the samplesheet in this situation ? Thanks !

Hi there, just wanted to manage expectations / clarify some things. WISP is the module that will be added to the next release and is specifically designed for tumor informed minimal residual disease monitoring (i.e. it is mostly to report on tumor fraction in ctDNA). for matched biopsies we run samples separately (with the same normal as you indicate) and then integrate data downstream of the pipeline. As described here: https://pubmed.ncbi.nlm.nih.gov/34373653/.

Hi, I have done the multiple tumors for one normal before by first mapping all of the normals (e.g. using

processes_manual: "alignment,redux"

and giving in a samplesheet only with normals) and then in a second round providing the bams for the normals and the fastqs for the tumors:

group_id,subject_id,sample_id,sample_type,sequence_type,filetype,filepath
PATIENT1,PATIENT1,PATIENT1-T,tumor,dna,bam_redux,<gs://bucket/PATIENT1-T.dna.redux.bam>
PATIENT1,PATIENT1,PATIENT1-T,tumor,dna,bai,<gs://bucket/PATIENT1-T.dna.redux.bam.bai>
PATIENT1,PATIENT1,PATIENT1-T,tumor,dna,redux_jitter_tsv,<gs://bucket/PATIENT1-T.dna.jitter_params.tsv>
PATIENT1,PATIENT1,PATIENT1-T,tumor,dna,redux_ms_tsv,<gs://bucket/PATIENT1-T.dna.ms_table.tsv.gz>

Just make sure your group_id, subject_id etc are unique for the multiple tumors, and specify the normal files for each of those tumors and not only once. It is not 100% ideal, but at least you avoid mapping the normal multiple times indeed.

Thanks to both of you ! Finally I modified the code to remove the verification and the error

germline analysis only is not supported

, to be able to align the normal alone with

processes-manual

. Otherwise the normals had to be paired to a tumor to run anything.