LogR correction, 10% overlap check

Hello,

I am trying to use ASCAT for a project that requires quick ploidy estimation and later for mutational signature analysis of pediatric whole genome sequencing data. I use a github install of your package, and have a local build working with seemingly correct output.

However, there is a check (see below) for LogR correction makes the program fail, and I am curious as to why this check for a 10% overlap is in place. In my data this overlap is around 8.5%. When I comment this line out in a local build the output seems accurate, but I would like to know what this error means. You will find details below, but I am mostly interested in the purpose of this check.

Thank you in advance! :)

Best, Maarten

The LogR correction check in question, happening here

Error in ascat.correctLogR(ascat.bc, GCcontentfile = "G1000/GC_G1000_hg38.txt",  :
  length(ovl) > nrow(ASCATobj$Tumor_LogR)/10 is not TRUE

R script calling ASCAT

tumor_bam_path = args[1]
normal_bam_path = args[2]
tumor_label = args[3]
normal_label = args[4] 
sex = args[5] # 'XX' or 'XY'

# Chromosome names
if(sex == 'XX'){chrom_names = c(1:22, "X")}
if(sex == 'XY'){chrom_names = c(1:22, "X", "Y")}

# ASCAT loading
library(devtools)
# devtools::install("USER/software/ascat/ASCAT")
library(ASCAT)

# Intermediate files:
tumor_logR = paste("output/", tumor_label, "_LogR.txt", sep = "")
tumor_BAF = paste("output/", tumor_label, "_BAF.txt", sep = "")
normal_logR = paste("output/", normal_label, "_logR.txt", sep = "")
normal_BAF = paste("output/", normal_label, "_BAF.txt", sep = "")

# ASCAT running
genomeVersion = 'hg38'
ascat.prepareHTS(
  tumourseqfile = tumor_bam_path,
  normalseqfile = normal_bam_path,
  tumourname = tumor_label,
  normalname = normal_label,
  allelecounter_exe = "USER/software/miniconda3/envs/ascat/bin/alleleCounter",
  alleles.prefix = "USER/allelecounter/G1000/G1000_alleles_hg38_chr",
  loci.prefix = "USER/allelecounter/G1000/G1000_loci_hg38_chr",
  skip_allele_counting_normal = T,
  skip_allele_counting_tumour = T,
  gender = sex,
  genomeVersion = genomeVersion,
  nthreads = 10,
  tumourLogR_file = tumor_logR,
  tumourBAF_file = tumor_BAF,
  normalLogR_file = normal_logR,
  normalBAF_file = normal_BAF,
  chrom_names = chrom_names
  )
ascat.bc = ascat.loadData(
  Tumor_LogR_file = tumor_logR, 
  Tumor_BAF_file = tumor_BAF, 
  Germline_LogR_file = normal_logR, 
  Germline_BAF_file = normal_BAF, 
  gender = sex, 
  genomeVersion = genomeVersion
  )
ascat.plotRawData(ascat.bc, img.prefix = "Before_correction_")
ascat.bc = ascat.correctLogR(
  ascat.bc, 
  GCcontentfile = "USER/allelecounter/G1000/GC_G1000_hg38.txt", 
  replictimingfile = "USER/allelecounter/G1000/RT_G1000_hg38.txt"
  )
ascat.plotRawData(ascat.bc, img.prefix = "After_correction_")
ascat.bc = ascat.aspcf(ascat.bc)
ascat.plotSegmentedData(ascat.bc)
ascat.output = ascat.runAscat(ascat.bc, gamma=1, write_segments = T)
QC = ascat.metrics(ascat.bc,ascat.output)
save(ascat.bc, ascat.output, QC, file = 'ASCAT_objects.Rdata')

Before running ASCAT, I run alleleCounter separately for each chromosome {1..22, X} using the command below:

alleleCounter
    -b USER/data/BAM/{bam_file} \
    -l USER/allelecounter/G1000/G1000_loci_hg38_chr{chr_num}.txt \
    -o USER/allelecounter/output/{sample_id}/{sample_id}_alleleFrequencies_chr{chr_num}.txt \
    -m 20 \
    -q 35 \
    --dense-snps

VanLoo-lab / ascat

LogR correction, 10% overlap check #180