analysis_BASIS-CFTR_assembly.Rmd

---
title: "Analysis for final CFTR BASIS assembly"
author: "Pierre Murat, Askar Kleefeldt"
date: "`r format(Sys.time(), '%d %B, %Y')`"
output: 
  html_document:
    toc: true
    toc_float: true
    toc_collapsed: true
    toc_depth: 3
    number_sections: true
    theme: lumen

---
# Prepare and index genomes
Human and bacterial genomes are catenated to generate a chimeric genomes used for alignment.

Human genome: Ensembl hg38/GRCh38
E. Coli: MDS42
Helper vectors: pKW20 and pLF118

```{bash, eval = F}
srun -c 112 --pty bash
# Extract contigs names
grep "^>" ./Genomes/hg38/Homo_sapiens.GRCh38.dna.primary_assembly.fa > ./BASIS/Sequences/contig_names/hg38.seqname.txt
grep "^>" ./BASIS/Sequences/MDS42_WT_old_seq_better_annotation.fa > ./BASIS/Sequences/contig_names/MDS42.seqname.txt
grep "^>" ./BASIS/Sequences/p20_helper-plasmid_BASIS.fa > ./BASIS/Sequences/contig_names/p20.seqname.txt
grep "^>" ./BASIS/Sequences/pLF118_helper_plasmid_BACs.fa > ./BASIS/Sequences/contig_names/pLF118.seqname.txt

# Manually modify headers of bacterial fa files / re-run
# Join genomes and indexes
cat ./Genomes/hg38/Homo_sapiens.GRCh38.dna.primary_assembly.fa \
    ./BASIS/Sequences/MDS42_WT_old_seq_better_annotation.fa \
    ./BASIS/Sequences/p20_helper-plasmid_BASIS.fa \
    ./BASIS/Sequences/pLF118_helper_plasmid_BACs.fa \
    > ./BASIS/Genomes/hg38.MDS42.fa
grep "^>" ./BASIS/Genomes/hg38.MDS42.fa > ./BASIS/Sequences/contig_names/hg38.MDS42.seqname.txt

./Software/bwa/bwa index ./BASIS/Genomes/hg38.MDS42.fa
samtools faidx ./BASIS/Genomes/hg38.MDS42.fa
./Software/gatk-4.2.5.0/gatk CreateSequenceDictionary -R ./BASIS/Genomes/hg38.MDS42.fa
```

# Create a reference haplotype
## Read alignment

FASTQ from individual input BACs (CFTR_BAC01, CFTR_BAC02, CFTR_BAC03) are combined and aligned to the reference genome.

```{bash, eval = F}

# align concatenated BAC fastq files to chimeric genome
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.fa CFTR_v02/Data/fastq_BACs/BACs.r1.fq.gz CFTR_v02/Data/fastq_BACs/BACs.r2.fq.gz > CFTR_v02/Data/SAM/BACs.hg38.MDS42.sam

# sort and generate BAM file
Software/samtools-1.16.1/samtools sort -@ 4 -m 2G CFTR_v02/Data/SAM/BACs.hg38.MDS42.sam -o CFTR_v02/Data/BAM/BACs.hg38.MDS42.bam

# index BAM file
Software/samtools-1.16.1/samtools index CFTR_v02/Data/BAM/BACs.hg38.MDS42.bam

# count number of reads
Software/samtools-1.16.1/samtools view -c CFTR_v02/Data/BAM/BACs.hg38.MDS42.bam
# 20,331,134 reads

# remove reads with multiple alignments
Software/samtools-1.16.1/samtools view -h CFTR_v02/Data/BAM/BACs.hg38.MDS42.bam | grep -v -e 'XA:Z:' -e 'SA:Z:' | Software/samtools-1.16.1/samtools view -b > CFTR_v02/Data/BAM/BACs.hg38.MDS42.unique.bam

# select properly paired reads
Software/samtools-1.16.1/samtools view -q 10 -F 1284 -f 0x02 -b CFTR_v02/Data/BAM/BACs.hg38.MDS42.unique.bam > CFTR_v02/Data/BAM/BACs.hg38.MDS42.paired.bam

# count number of reads
Software/samtools-1.16.1/samtools view -c CFTR_v02/Data/BAM/BACs.hg38.MDS42.paired.bam
# 15,520,500 reads

# Dump bacterial chromosomes
# Create an index file
awk '/^[0-9]*\t/ {printf("%s\t0\t%s\n",$1,$2);}' Genomes/hg38/Homo_sapiens.GRCh38.dna.primary_assembly.fa.fai > CFTR_v02/Data/BAM/hum.index.bed

# filter only human chromosomes
Software/samtools-1.16.1/samtools view -L CFTR_v02/Data/BAM/hum.index.bed -o CFTR_v02/Data/BAM/BACs.only.hum.bam CFTR_v02/Data/BAM/BACs.hg38.MDS42.paired.bam

# Compute coverage using deeptool bam coverage
Software/miniconda3/envs/deeptools/bin/bamCoverage --bam CFTR_v02/Data/BAM/BACs.only.hum.bam -o CFTR_v02/Data/BW/BACs.only.hum.50nt.bw -of bigwig --binSize 50 --normalizeUsing CPM
```


## Base quality score recalibration (BCSR)
In order to call variants confidently, we need to recalibrate the base quality scores using an established pipeline available from GATK. Candidate “germline” short variants are then filtered by Variant Quality Score Recalibration (VQSR).

```{bash, eval = F}
# Mark duplicates and sort BAM
Software/gatk-4.3.0.0/gatk MarkDuplicates -I CFTR_v02/Data/BAM/BACs.only.hum.bam -O CFTR_v02/Data/BAMdup/BACs.only.hum.dup.bam -M CFTR_v02/Data/BAMdup/BACs.dup.metrics.txt

Software/gatk-4.3.0.0/gatk SortSam -I CFTR_v02/Data/BAMdup/BACs.only.hum.dup.bam -O CFTR_v02/Data/BAMdup/BACs.only.hum.dup.sort.bam -SO coordinate

# Add read groups (needed for picard/BQSR)
Software/gatk-4.3.0.0/gatk AddOrReplaceReadGroups -I CFTR_v02/Data/BAMdup/BACs.only.hum.dup.sort.bam -O CFTR_v02/Data/BAMdup/BACs.only.hum.dup.sort.group.bam -LB BAC_library -PL ILLUMINA -PU unknown -SM Normal

# Base quality score recalibration (BQSR)
# Generate recalibration table for BQSR
Software/gatk-4.3.0.0/gatk BaseRecalibrator -I CFTR_v02/Data/BAMdup/BACs.only.hum.dup.sort.group.bam -R Genomes/hg38/Homo_sapiens.GRCh38.dna.primary_assembly.fa --known-sites ./ChromTrans/Dataset/GATK_resource/BQSR.known.sites.hg38.recode.sort.vcf -O CFTR_v02/Data/BQSR/BACs.only.hum.recal.data.table

# Apply BQSR
Software/gatk-4.3.0.0/gatk ApplyBQSR -R Genomes/hg38/Homo_sapiens.GRCh38.dna.primary_assembly.fa -I CFTR_v02/Data/BAMdup/BACs.only.hum.dup.sort.group.bam --bqsr-recal-file CFTR_v02/Data/BQSR/BACs.only.hum.recal.data.table -O CFTR_v02/Data/BQSR/BACs.only.hum.bqsr.bam

Software/samtools-1.16.1/samtools index CFTR_v02/Data/BQSR/BACs.only.hum.bqsr.bam

```

## Define haplotype
Identify germline mutations with HaplotypeCaller and filtered using Variant Quality Score Recalibration (VQSR). VQSR uses an adaptive error model trained on “true sites” provided as input, typically HapMap and Omni 2.5M SNP chip array sites. Input are available from the GATK Resource Bundle, https://console.cloud.google.com/storage/browser/genomics-public-data/resources/broad/hg38/v0/ but need to be reformated to be used with Ensembl genomes.

```{bash, eval = F}
# Call variants
Software/gatk-4.3.0.0/gatk HaplotypeCaller -R Genomes/hg38/Homo_sapiens.GRCh38.dna.primary_assembly.fa -I CFTR_v02/Data/BQSR/BACs.only.hum.bqsr.bam -O CFTR_v02/Data/VCF/HAPLO/BACs.haplotype.vcf

# Build a recalibration model with VariantRecalibrator
Software/gatk-4.3.0.0/gatk VariantRecalibrator -R ./Genomes/hg38/Homo_sapiens.GRCh38.dna.primary_assembly.fa -V CFTR_v02/Data/VCF/HAPLO/BACs.haplotype.vcf -AS --resource:hapmap,known=false,training=true,truth=true,prior=15.0 ./ChromTrans/Dataset/GATK_resource/hapmap_3.3.hg38.sites.recode.sort.vcf --resource:omni,known=false,training=true,truth=true,prior=12.0 ./ChromTrans/Dataset/GATK_resource/1000G_omni2.5.hg38.recode.sort.vcf --resource:1000G,known=false,training=true,truth=false,prior=10.0 ./ChromTrans/Dataset/GATK_resource/1000G_phase1.snps.high_confidence.hg38.recode.sort.vcf --resource:dbsnp,known=true,training=false,truth=false,prior=2.0 ./ChromTrans/Dataset/GATK_resource/Homo_sapiens_assembly38.dbsnp138.recode.sort.vcf -an QD -an MQ -an MQRankSum -an ReadPosRankSum -an FS -an SOR -mode SNP --max-gaussians 4 -O CFTR_v02/Data/VCF/HAPLO/BACs.output.AS.recal --tranches-file CFTR_v02/Data/VCF/HAPLO/BACs.output.AS.tranches --rscript-file CFTR_v02/Data/VCF/HAPLO/BACs.output.plots.AS.R

# Apply recalibration
Software/gatk-4.3.0.0/gatk ApplyVQSR -R ./Genomes/hg38/Homo_sapiens.GRCh38.dna.primary_assembly.fa -V CFTR_v02/Data/VCF/HAPLO/BACs.haplotype.vcf -O CFTR_v02/Data/VCF/HAPLO/BACs.haplotype.VQSR.vcf -AS --truth-sensitivity-filter-level 99.0 --tranches-file CFTR_v02/Data/VCF/HAPLO/BACs.output.AS.tranches --recal-file CFTR_v02/Data/VCF/HAPLO/BACs.output.AS.recal -mode SNP

```

## Generate an alternative reference sequence
The previous identify germline variants are used to generate an alternative reference genome (FastaAlternateReferenceMaker). We first select variants on the regions covered by the BACs. To do so we look at region of chr7 with sufficient coverage to support domains covered by BACs.

```{R, eval = F}
library(rtracklayer)
BAC.cov.gr <- import.bw("./CFTR_v02/Data/BW/BACs.only.hum.50nt.bw")
BAC.cov.chr7.gr <- BAC.cov.gr[seqnames(BAC.cov.gr) == "7"]

# Identify domains with more than 10 CPM per 50 nt and add an extra 10 kb both side
BAC.domain.start <- min(as.data.frame(BAC.cov.chr7.gr[BAC.cov.chr7.gr$score >= 10])$start) - 10000
BAC.domain.end <- max(as.data.frame(BAC.cov.chr7.gr[BAC.cov.chr7.gr$score >= 10])$end) + 10000 

```

Replace the reference bases at variation sites for the interval covered by the BAC library
WARNING: FastaAlternateReferenceMaker alters the name of the chromosomes, so we need to swap the contigs name

```{bash, eval = F}
# alternative reference maker with full genome
# alternative reference maker gave error since it is missing a fa.dict file

# made dict file using Picard and CreateSequenceDictionary
Software/gatk-4.3.0.0/gatk CreateSequenceDictionary -R CFTR_v02/Genomes/hg38.MDS42.fa -O CFTR_v02/Genomes/hg38.MDS42.fa.dict

# need to rename the dict file such that it does not end on .fa.dict but just on .dict
# this was liukely changed manually in the first instance already

# rerun alternate reference maker
Software/gatk-4.3.0.0/gatk FastaAlternateReferenceMaker -R CFTR_v02/Genomes/hg38.MDS42.fa -O CFTR_v02/Genomes/hg38.MDS42.alternate.BAC.fa -V CFTR_v02/Data/VCF/HAPLO/BACs.haplotype.VQSR.vcf

grep "^>" CFTR_v02/Genomes/hg38.MDS42.alternate.BAC.fa > CFTR_v02/Sequences/contig_names/hg38.MDS42.alternate.seqname.txt

```

Rename fasta headers using a lookup table with seqkit (https://bioinf.shenwei.me/seqkit/)

```{bash, eval = F}
# installed seqkit via conda
Software/miniconda3/bin/conda install -c bioconda seqkit

sed '/^>/ s/ .*//' CFTR_v02/Genomes/hg38.MDS42.alternate.BAC.fa > CFTR_v02/Genomes/temp.fa
grep "^>" CFTR_v02/Genomes/temp.fa > CFTR_v02/Sequences/contig_names/temp.seqname.txt

```


```{R, eval = F}
old.name <- read.table("./CFTR_v02/Sequences/contig_names/temp.seqname.txt", sep = ",")
new.name <- read.table("./CFTR_v02/Sequences/contig_names/hg38.MDS42.seqname.txt", sep = ",")

corr.df <- cbind.data.frame(gsub(">", "", old.name$V1), gsub(">", "", new.name$V1))
write.table(corr.df, "./CFTR_v02/Sequences/contig_names/corr.table",
            sep = "\t", quote = F, col.names = F, row.names = F)
```

```{bash, eval = F}
# Use seqkit to replace the correct names
Software/miniconda3/bin/seqkit replace -p '([0-9]+)' -r '{kv}$2' -k CFTR_v02/Sequences/contig_names/corr.table CFTR_v02/Genomes/temp.fa > CFTR_v02/Genomes/hg38.MDS42.alternate.fa

grep "^>" CFTR_v02/Genomes/hg38.MDS42.alternate.fa > CFTR_v02/Sequences/contig_names/hg38.MDS42.alternate.seqname.txt

# Skipped removal of temporary files

# Index alternate genome
Software/bwa/bwa index CFTR_v02/Genomes/hg38.MDS42.alternate.fa

```


# Somatic variant calling
## Align BACs and BASIS samples to alternate chromosome

```{bash, eval = F}

# pool of BACs
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.fa CFTR_v02/Data/fastq_BACs/BACs.r1.fq.gz CFTR_v02/Data/fastq_BACs/BACs.r2.fq.gz > CFTR_v02/Data/SAM/BACs.hg38.MDS42.alternate.sam

# individual BACs
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.fa CFTR_v02/Data/fastq_BACs/BAC01/11A11_S11_R1_001.fastq.gz CFTR_v02/Data/fastq_BACs/BAC01/11A11_S11_R2_001.fastq.gz > CFTR_v02/Data/SAM/BAC01.hg38.MDS42.alternate.sam

Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.fa CFTR_v02/Data/fastq_BACs/BAC02/12A12_S12_R1_001.fastq.gz CFTR_v02/Data/fastq_BACs/BAC02/12A12_S12_R2_001.fastq.gz > CFTR_v02/Data/SAM/BAC02.hg38.MDS42.alternate.sam

Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.fa CFTR_v02/Data/fastq_BACs/BAC03/13A13_S13_R1_001.fastq.gz CFTR_v02/Data/fastq_BACs/BAC03/13A13_S13_R2_001.fastq.gz > CFTR_v02/Data/SAM/BAC03.hg38.MDS42.alternate.sam

# CFTR samples
# intermediate step
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.fa CFTR_v02/Data/fastq_CFTR_intermediate/14A14_S14_R1_001.fastq.gz CFTR_v02/Data/fastq_CFTR_intermediate/14A14_S14_R2_001.fastq.gz > CFTR_v02/Data/SAM/CFTRintermediate.hg38.MDS42.alternate.sam

# final step
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.fa CFTR_v02/Data/fastq_CFTR_cl3/6A6_S6_R1_001.fastq.gz CFTR_v02/Data/fastq_CFTR_cl3/6A6_S6_R2_001.fastq.gz > CFTR_v02/Data/SAM/CFTRfinal.hg38.MDS42.alternate.sam

```

## Process all aligned data
Data are then processed as previously using /BASIS/Scripts/Read_processing_SLURM.sh script. Shell script is reported below:

```{bash, eval = F}
#!/bin/bash -l
for SAM in ./CFTR_v02/Data/SAM/*.hg38.MDS42.alternate.sam
do
# Recover file names
FILE=${SAM%.hg38.MDS42.alternate.sam}
SAMPLE=${FILE##*/}
echo $SAMPLE
# Sort and generate BAM file
./Software/samtools-1.16.1/samtools sort -@ 4 -m 2G ./CFTR_v02/Data/SAM/$SAMPLE.hg38.MDS42.alternate.sam -o ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.bam

# Index BAM file
./Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.bam

# Generate stats
./Software/samtools-1.16.1/samtools idxstats ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.bam > ./CFTR_v02/Data/IDXSTATS/$SAMPLE.hg38.MDS42.alternate.idxstats.txt

# Remove reads with multiple alignments
./Software/samtools-1.16.1/samtools view -h ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.bam | grep -v -e 'XA:Z:' -e 'SA:Z:' | ./Software/samtools-1.16.1/samtools view -b > ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.unique.bam

# Select properly paired reads
./Software/samtools-1.16.1/samtools view -q 10 -F 1284 -f 0x02 -b ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.unique.bam > ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.paired.bam
./Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.paired.bam

# Recompute stats
./Software/samtools-1.16.1/samtools idxstats ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.paired.bam > ./CFTR_v02/Data/IDXSTATS/$SAMPLE.hg38.MDS42.alternate.paired.idxstats.txt

# Dump bacterial chromosomes
./Software/samtools-1.16.1/samtools view -L ./CFTR_v02/Data/BAM/hum.index.bed \
                                                          -o ./CFTR_v02/Data/BAM/$SAMPLE.alternate.only.hum.bam \
                                                             ./CFTR_v02/Data/BAM/$SAMPLE.hg38.MDS42.alternate.paired.bam 

./Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAM/$SAMPLE.alternate.only.hum.bam

# Compute coverage using deeptool bam coverage
./Software/miniconda3/envs/deeptools/bin/bamCoverage \
      --bam ./CFTR_v02/Data/BAM/$SAMPLE.alternate.only.hum.bam \
      -o    ./CFTR_v02/Data/BW/$SAMPLE.alternate.only.hum.50nt.bw \
      -of bigwig \
      --binSize 50 \
      --normalizeUsing CPM

done
```

To extract a specific region from a fasta file, samtools faidx can be used. Here used to extract the target CFTR region to understand whether the sequences substantially differ.

```{bash, eval = F}
Software/samtools-1.16.1/samtools faidx CFTR_v02/Genomes/hg38.MDS42.alternate.fa 7:117450000-117700000 > CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa
Software/samtools-1.16.1/samtools faidx CFTR_v02/Genomes/hg38.MDS42.fa 7:117450000-117700000 > CFTR_v02/Genomes/hg38.MDS42.chr7_CFTR.fa


```

## Mark duplicates and group reads
Data are then processed as previously using the ./BASIS/Scripts/Picard_mark_SLURM.sh script. Shell script is reported below:

```{bash, eval = F}
#!/bin/bash -l
for BAM in ./CFTR_v02/Data/BAM/*.alternate.only.hum.bam
do
# Recover file names
FILE=${BAM%.alternate.only.hum.bam}
SAMPLE=${FILE##*/}
echo $SAMPLE
# Mark duplicates and sort BAM
./Software/gatk-4.3.0.0/gatk MarkDuplicates \
    -I ./CFTR_v02/Data/BAM/$SAMPLE.alternate.only.hum.bam \
    -O ./CFTR_v02/Data/BAMdup/$SAMPLE.alternate.only.hum.dup.bam \
    -M ./CFTR_v02/Data/BAMdup/$SAMPLE.dup.metrics.txt
./Software/gatk-4.3.0.0/gatk SortSam \
    -I ./CFTR_v02/Data/BAMdup/$SAMPLE.alternate.only.hum.dup.bam \
    -O ./CFTR_v02/Data/BAMdup/$SAMPLE.alternate.only.hum.dup.sort.bam \
    -SO coordinate
# Add read groups (needed for Mutect2)
./Software/gatk-4.3.0.0/gatk AddOrReplaceReadGroups \
    -I ./CFTR_v02/Data/BAMdup/$SAMPLE.alternate.only.hum.dup.sort.bam \
    -O ./CFTR_v02/Data/BAMdup/$SAMPLE.alternate.only.hum.dup.sort.group.bam \
    -LB $SAMPLE \
    -PL ILLUMINA \
    -PU unknown \
    -SM $SAMPLE
done
```

## Candidate short variants calling
Candidate short variants are called with Mutect2 with the tumor (= CFTR samples) with matched normal (= BACs).

```{bash, eval = F}
# Create a sequence dictionary file for the alternative genome
Software/gatk-4.3.0.0/gatk CreateSequenceDictionary -R ./CFTR_v02/Genomes/hg38.MDS42.alternate.fa

# Check read group information
Software/samtools-1.16.1/samtools view -H ./CFTR_v02/Data/BAMdup/BACs.alternate.only.hum.dup.sort.group.bam | grep '^@RG'
# @RG	ID:1	LB:BACs	PL:ILLUMINA	SM:BACs	PU:unknown

Software/samtools-1.16.1/samtools view -H ./CFTR_v02/Data/BAMdup/CFTRfinal.alternate.only.hum.dup.sort.group.bam | grep '^@RG'
# @RG	ID:1	LB:CFTRfinal	PL:ILLUMINA	SM:CFTRfinal	PU:unknown

# Index BAM files
Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAMdup/BACs.alternate.only.hum.dup.sort.group.bam 
Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAMdup/CFTRintermediate.alternate.only.hum.dup.sort.group.bam
Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAMdup/CFTRfinal.alternate.only.hum.dup.sort.group.bam

# Run Mutect2 (Tumor with matched normal mode)
# CFTR samples intermediate and final assembly
Software/gatk-4.3.0.0/gatk Mutect2 -R ./CFTR_v02/Genomes/hg38.MDS42.alternate.fa -I ./CFTR_v02/Data/BAMdup/BACs.alternate.only.hum.dup.sort.group.bam -I ./CFTR_v02/Data/BAMdup/CFTRintermediate.alternate.only.hum.dup.sort.group.bam -normal BACs -O ./CFTR_v02/Data/VCF/SOMA/CFTRintermediate.mutect2.vcf

Software/gatk-4.3.0.0/gatk Mutect2 -R ./CFTR_v02/Genomes/hg38.MDS42.alternate.fa -I ./CFTR_v02/Data/BAMdup/BACs.alternate.only.hum.dup.sort.group.bam -I ./CFTR_v02/Data/BAMdup/CFTRfinal.alternate.only.hum.dup.sort.group.bam -normal BACs -O ./CFTR_v02/Data/VCF/SOMA/CFTRfinal.mutect2.vcf

```

## Filter short variants
Candidate short variants are then filtered using the shell script /BASIS/Scripts/Mutect_filter_SLURM.sh and /BASIS/Scripts/Mutect_filter_SLURM.2.sh, reported below:

```{bash, eval = F}
#!/bin/bash -l
for VCF in ./CFTR_v02/Data/VCF/SOMA/*.mutect2.vcf
do
# Recover file names
FILE=${VCF%.mutect2.vcf}
SAMPLE=${FILE##*/}
echo $SAMPLE
# Filter Mutect2 VCF
./Software/gatk-4.3.0.0/gatk FilterMutectCalls \
   -V ./CFTR_v02/Data/VCF/SOMA/$SAMPLE.mutect2.vcf \
   -R ./CFTR_v02/Genomes/hg38.MDS42.alternate.fa \
   -O ./CFTR_v02/Data/VCF/SOMA/$SAMPLE.mutect2.filter.vcf
# Generate table from VCF
./Software/gatk-4.3.0.0/gatk VariantsToTable \
     -V ./CFTR_v02/Data/VCF/SOMA/$SAMPLE.mutect2.filter.vcf \
     -F CHROM -F POS -F REF -F ALT -F FILTER -GF AF -GF GT \
     -O ./CFTR_v02/Data/VCF/SOMA/$SAMPLE.mutect2.filter.tsv
done
```

## Mutation analysis / curation

```{R, eval = F}
#R
library(dplyr)
# Load all detected somatic mutations
BASIS.files <- list.files(path = "./CFTR_v02/Data/VCF/SOMA", pattern = "\\.mutect2.filter.tsv")
BASIS.samples <- gsub(".mutect2.filter.tsv", "", BASIS.files)
BASIS.path <- paste("./CFTR_v02/Data/VCF/SOMA", BASIS.files, sep = "/")
BASIS.vcf <- tibble()
for (i in 1:length(BASIS.path)) {
  print(BASIS.samples[i])
  vcf.i <- read.table(BASIS.path[i], header = T) %>% mutate(SAMPLE = BASIS.samples[i])
  colnames(vcf.i) <- c("CHROM", "POS", "REF", "ALT", "FILTER", "BAC.AF", "BAC.GT", "BASIS.AF", "BASIS.GT", "SAMPLE")
  BASIS.vcf <- rbind(BASIS.vcf, vcf.i)
}
BASIS.vcf <- BASIS.vcf %>% filter(CHROM == 7 & POS >= 117465000 & POS <= 117680000)
BASIS.ligth.vcf <- BASIS.vcf %>% filter(BASIS.AF >= 0.25) %>% dplyr::select(CHROM, POS, REF, ALT) %>% unique() %>% filter(REF != "TRUE")
BASIS.true.vcf <- BASIS.vcf %>% filter(POS %in% BASIS.ligth.vcf$POS) %>% arrange(POS)
# Save vcf summary
write.csv(BASIS.true.vcf, "./CFTR_v02/Data/VCF/SOMA.summary.csv")
# The resulting file is manually modified to report additional information
```

No mutations detected after running pipeline.

# Final verification and alignment of CFTR BACs for data visualisation

```{bash, eval = F}

# index genome
Software/bwa/bwa index ./CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa
Software/samtools-1.16.1/samtools faidx ./CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa

# pool of BACs
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa CFTR_v02/Data/fastq_BACs/BACs.r1.fq.gz CFTR_v02/Data/fastq_BACs/BACs.r2.fq.gz > CFTR_v02/Data/SAM/CFTR_only/BACs.hg38.MDS42.alternate.chr7_CFTR.sam

# individual BACs
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa CFTR_v02/Data/fastq_BACs/BAC01/11A11_S11_R1_001.fastq.gz CFTR_v02/Data/fastq_BACs/BAC01/11A11_S11_R2_001.fastq.gz > CFTR_v02/Data/SAM/CFTR_only/BAC01.hg38.MDS42.alternate.chr7_CFTR.sam

Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa CFTR_v02/Data/fastq_BACs/BAC02/12A12_S12_R1_001.fastq.gz CFTR_v02/Data/fastq_BACs/BAC02/12A12_S12_R2_001.fastq.gz > CFTR_v02/Data/SAM/CFTR_only/BAC02.hg38.MDS42.alternate.chr7_CFTR.sam

Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa CFTR_v02/Data/fastq_BACs/BAC03/13A13_S13_R1_001.fastq.gz CFTR_v02/Data/fastq_BACs/BAC03/13A13_S13_R2_001.fastq.gz > CFTR_v02/Data/SAM/CFTR_only/BAC03.hg38.MDS42.alternate.chr7_CFTR.sam

# CFTR samples
# intermediate step
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa CFTR_v02/Data/fastq_CFTR_intermediate/14A14_S14_R1_001.fastq.gz CFTR_v02/Data/fastq_CFTR_intermediate/14A14_S14_R2_001.fastq.gz > CFTR_v02/Data/SAM/CFTR_only/CFTRintermediate.hg38.MDS42.alternate.chr7_CFTR.sam

# final step
Software/bwa/bwa mem -M -t 7 CFTR_v02/Genomes/hg38.MDS42.alternate.chr7_CFTR.fa CFTR_v02/Data/fastq_CFTR_cl3/6A6_S6_R1_001.fastq.gz CFTR_v02/Data/fastq_CFTR_cl3/6A6_S6_R2_001.fastq.gz > CFTR_v02/Data/SAM/CFTR_only/CFTRfinal.hg38.MDS42.alternate.chr7_CFTR.sam

```

Process reads with similar script as done earlier.

```{bash, eval = F}
#!/bin/bash -l
for SAM in ./CFTR_v02/Data/SAM/CFTR_only/*.hg38.MDS42.alternate.chr7_CFTR.sam
do
# Recover file names
FILE=${SAM%.hg38.MDS42.alternate.chr7_CFTR.sam}
SAMPLE=${FILE##*/}
echo $SAMPLE
# Sort and generate BAM file
./Software/samtools-1.16.1/samtools sort -@ 4 -m 2G ./CFTR_v02/Data/SAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.sam -o ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.bam

# Index BAM file
./Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.bam

# Generate stats
./Software/samtools-1.16.1/samtools idxstats ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.bam > ./CFTR_v02/Data/IDXSTATS/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.idxstats.txt

# Remove reads with multiple alignments
./Software/samtools-1.16.1/samtools view -h ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.bam | grep -v -e 'XA:Z:' -e 'SA:Z:' | ./Software/samtools-1.16.1/samtools view -b > ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.unique.bam

# Select properly paired reads
./Software/samtools-1.16.1/samtools view -q 10 -F 1284 -f 0x02 -b ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.unique.bam > ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.paired.bam
./Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.paired.bam

# Recompute stats
./Software/samtools-1.16.1/samtools idxstats ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.paired.bam > ./CFTR_v02/Data/IDXSTATS/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.paired.idxstats.txt

# Dump bacterial chromosomes
./Software/samtools-1.16.1/samtools view -L ./CFTR_v02/Data/BAM/CFTR_only/hum.index.bed \
                                                          -o ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.alternate.chr7_CFTR.only.hum.bam \
                                                             ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.hg38.MDS42.alternate.chr7_CFTR.paired.bam 

./Software/samtools-1.16.1/samtools index ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.alternate.chr7_CFTR.only.hum.bam

# Compute coverage using deeptool bam coverage
./Software/miniconda3/envs/deeptools/bin/bamCoverage \
      --bam ./CFTR_v02/Data/BAM/CFTR_only/$SAMPLE.alternate.chr7_CFTR.only.hum.bam \
      -o    ./CFTR_v02/Data/BW/CFTR_only/$SAMPLE.alternate.chr7_CFTR.only.hum.50nt.bw \
      -of bigwig \
      --binSize 50 \
      --normalizeUsing CPM

done
```