# A short script used to quality check and filter the VCF file for 17-VLT002652

# Load the VCF file and define thresholds and ratios for filtering
vcf_file <- "~/Path/to/project/folder/17VLT2652_var_called.vcf"
vcf <- read.delim(vcf_file, skip = 29, stringsAsFactors = FALSE)
min_QUAL <- 100
min_frac_qual <- 0.1

# Filter on read quality
vcf_alt_1 <- subset(vcf, vcf$QUAL > min_QUAL)

okrow <- rep(c(1), nrow(vcf_alt_1))

# Filter out samples with too similar likelihood values for each allele
for (count in 1:nrow(vcf_alt_1)) {
  for (r in 10:ncol(vcf_alt_1)) {
    p <- as.numeric(strsplit(strsplit(vcf_alt_1[count,r],":")[[1]][2], ",")[[1]])
    if (as.numeric(max(p))>=25 && as.numeric(min(p))<min_frac_qual*as.numeric(max(p))){
      okrow[count] <- 1
    } else {
      okrow[count] <- 0
    }
  }
  #print(count)
}

vcf_temp_1 <- subset(cbind(vcf_alt_1, okrow), okrow==1)
vcf_alt_2 <- vcf_temp_1[1:ncol(vcf_temp_1)-1]

# Write out the new VCF file (without a header)
write.table(vcf_alt_2, file="~/Path/to/project/folder/17VLT2652_var_qual_checked_test.vcf", quote = FALSE, sep = '\t', row.names = FALSE)

################################################################################################################

library(vcfR)
library(ape)

# Load the VCF, genomic FASTA and GFF3 files for 17-VLT002652
vcf_file <- "~/Path/to/project/folder/17VLT2652_var_qual_checked_test.vcf"
dna_file <- "~/Path/to/project/folder/GCF_000006945.2_ASM694v2_genomic.fna"
gff_file <- "~/Path/to/project/folder/GCF_000006945.2_ASM694v2_genomic.gff"

vcf <- read.vcfR(vcf_file, verbose = FALSE)
dna <- ape::read.dna(dna_file, format = "fasta")
gff <- read.table(gff_file, sep="\t", quote="")

# Load the files into a chromR object
chrom <- create.chromR(name='S_Typhimurium_LT2', vcf=vcf, seq=dna, ann = gff)

# Plot the distributions of read depth, mapping quality and read quality to decide thresholds for filtering
plot(chrom)

# Filter based on the previously decided thresholds and plot again to see the difference
chrom_mask1 <- masker(chrom, min_QUAL = 100, max_DP = 400, min_MQ = 20)
plot(chrom_mask1)

# Compute the variant count per 1 Kb window and plot again
chrom_proc <- proc.chromR(chrom_mask1, verbose = TRUE)
plot(chrom_proc)

# Plot a dot plot to see the distribution of read depth and quality over the genome
chromoqc(chrom_proc, dp.alpha = 20)

# Write out the filtered VCF file
write.vcf(chrom_proc, file = "~/Path/to/project/folder/17VLT2652_var_qual_checked2_test.vcf.gz", mask = T)