ColorectalCancerMetastasis

---

Scripts for processing WES data from Colorectal Cancer Metastasis

Prerequisite

Alignment and Bam processing

Alignment

copy "align.sh" to working directory

for name in `ls -d *[0]`
do
  dna_analysis.sh -n  $name  -t  16  -s fastq -e reCal  -1  $fq1  -2  $fq2
done

Filters

copy "filter.sh" to mutect2/filteredVcf directory

for i in ../rawVcf/*.vcf.gz
do
  name=`basename $i .vcf.gz`
  ./filter.sh $name $i ../../contamination/$name.calculatecontamination.table  ../../ffpe/$name.artifact.pre_adapter_detail_metrics
done

Retrieve the selected variants

Get the bed file for each patient

for  i in BXY  DDY  DWY  LDS  WLF  YHD
do
  cd $i
  for j in *.vcf
  do
       convert2annovar.pl --format vcf4old  $j 2>/dev/null |cut -f 1-5  >>$i.raw.bed
  done
  sort -V $i.raw.bed|uniq >$i.sorted.uniq.bed
  cd ..
  done

Call gVcfs

 for i in */*.somatic_hardfiltered.clean.vcf
 do
 nohup gatk HaplotypeCaller -L `dirname $i`/`dirname $i`.sorted.uniq.bed --reference /database/ref/hg19.fa  --input ../bamFiles/`basename $i .somatic_hardfiltered.clean.vcf`.sorted.dedup.rg.recal.bam   --output `dirname $i`/`basename $i .somatic_hardfiltered.clean.vcf`.g.vcf   -ERC GVCF   --output-mode EMIT_ALL_SITES >log.`basename  $i .somatic_hardfiltered.clean.vcf` &
 done

Process the output gVcf

for i in BXY  DDY  DWY  LDS  WLF  YHD
do
   for i in $i/*.g.vcf;do echo " -V "$i;done|xargs  gatk CombineGVCFs -R /database/ref/hg19.fa  -O $i/${i}.g.vcf.gz
   gatk GenotypeGVCFs -L ${i}/${i}.sorted.uniq.bed  -R /database/ref/hg19.fa -V ${i}/${i}.g.vcf.gz  -O ${i}/${i}.jointlycall.vcf.gz
done

annotation + treeomics Input file prepare

for i in  BXY DDY  DWY  LDS  WLF  YHD
do
  table_annovar.pl ${i}/${i}.jointlycall.vcf.gz /database/annotation/annovar/humandb/ -buildver hg19 -out ${i}/${i} -remove -protocol refGene,cytoBand -operation g,r -nastring . -vcfinput
  gatk VariantsToTable -V ${i}/${i}.hg19_multianno.vcf  -F cytoBand -F POS -F REF -F ALT -F Func.refGene -F Gene.refGene -GF AD  -O ${i}/${i}.table
  awk 'BEGIN{OFS="\t"} {if(NR==1){printf "Chromosome\tPosition\tChange\tGene";for(i=7;i<=NF;i++){split($i,m,".");printf "\t"m[1]}printf "\n"}else{  if ($1 != "." && ($5 ~/^exonic$/ || $5 ~/^splicing$/ ) && $NF ~ /,0$/ ){printf "chr"$1"\t"$2"\t"$3">"$4"\t"$6;for(i=7;i<=NF;i++){split($i,m,",");printf "\t"m[2]}printf "\n"}}}' ${i}/${i}.table  >${i}/${i}.mutatedRead.txt
  awk 'BEGIN{OFS="\t"} {if(NR==1){printf "Chromosome\tPosition\tChange\tGene";for(i=7;i<=NF;i++){split($i,m,".");printf "\t"m[1]}printf "\n"}else{  if ($1 != "." && ($5 ~/^exonic$/ || $5 ~/^splicing$/ ) && $NF ~ /,0$/ ){printf "chr"$1"\t"$2"\t"$3">"$4"\t"$6;for(i=7;i<=NF;i++){split($i,m,",");printf "\t"(m[1]+m[2])}printf "\n"}}}' ${i}/${i}.table  >${i}/${i}.coverage.txt
  cp $i/$i.coverage.txt ~/treeomics/src/input/$i/
  cp $i/$i.mutatedRead.txt ~/treeomics/src/input/$i/
done

PyClone input preparation

Should run the initial two steps of treeomics input preparation to get table files

Split the merged table file into individual files

ONLY exonic and splicing variants reatined, sites on chrM or with alterantive reads in NORMAL sample were removed

for i in BXY DDY  DWY  LDS  WLF  YHD
do
awk 'BEGIN{OFS="\t"}{ if(NR==1){for(i=7;i<=NF;i++){split($i,n,".");f[i]=n[1]; printf "" >f[i]".txt"}}  else{ split($1,h,"[pq]"); for(i=7;i<=NF;i++){   split($i,m,",");  if($1 != "." && ($5 ~/^exonic$/ || $5 ~/^splicing$/) && length(m)==2  && $NF ~ /,0$/ ){printf "chr"h[1]"\t"($2-1)"\t"$2"\tchr"h[1]":"$2":"$3":"$4"\t"m[1]"\t"m[2]"\t2\t0\t-""\n"  >>f[i]".txt"   }     }  }  }' $i.table
done

Get the copy number info

for i in *.seg
do
  # the copy number MUST be integrate
  intersectBed -a `basename $i .seg`.txt -b  $i -wb |awk 'BEGIN{OFS="\t"}{if(NR==1){print  "mutation_id\tref_counts\tvar_counts\tnormal_cn\tminor_cn\tmajor_cn\n"$4,$5,$6,$7,$8,int(2^$14 * 2 +0.5)}else{print $4,$5,$6,$7,$8,int(2^$14 * 2 +0.5)}}' - >`basename $i .seg`.tsv
done
  # Missing normal files in the above calculation, create tsv files for normal samples
for i in *NOM*.txt
do
    awk 'BEGIN{OFS="\t"}{print $4,$5,$6,$7,$8,2}' $i |sed "1imutation_id\tref_counts\tvar_counts\tnormal_cn\tminor_cn\tmajor_cn">`basename $i .txt`.tsv
done

Run the Pyclone

Create the config file maunally according the demo

  # activate the vitrual python enviorment
  source activate python27
  # build the yaml files
  for i in ../tsv/*.tsv
  do
      PyClone build_mutations_file --prior total_copy_number --in_file $i --out_file `basename $i .tsv`.yaml
  done
  # run the main program of PyClone
  for i in bxy ddy dwy lds wlf yhd
  do
      mkdir trace_${i}
      PyClone run_analysis --config_file config.${i}.yaml --outfile trace_${i}   --seed 1986
      PyClone build_table  --config_file config.${i}.yaml --outfile ${i}.table --table_type loci
  done

Create gene-based tsv files

Use the tsv files created above; Use homemake script intersect

  for i in *.tsv
  do
    name=`basename $i .tsv`
    # Prepare the annotation input files
    tail -n +2 $i |awk -F "[:\t]" 'BEGIN{OFS="\t"}{print $1,$2-length($3)+1,$2,$3,$4,$0}' - >$name.avinput
    # Annotation
    annotate_variation.pl -buildver hg19 -outfile $name $name.avinput /database/annotation/annovar/humandb
    # Remove synonymous variants
    awk '$2~/^synonymous/' $name.exonic_variant_function >$name.tmp
    intersect  -c 8 9  -r  $name.variant_function -f $name.variant_function $name.tmp -o 1 >$name.txt
    rm $name.tmp
    # Create tsv files (should create tsv folder in case overlap original tsv)
    mkdir tsv
    sort -k2  $name.txt |awk 'BEGIN{OFS="\t"}{if($2~/\(/){sub(/\(\S+\)/,"",$2)};m[$2]=m[$2]+$9;n[$2]=n[$2]+$10;h[$2]=h[$2]+$13;a[$2]=a[$2]+1}END{for(k in m){print  k,m[k],n[k],2,0,int(h[k]/a[k])}}' -  >tsv/$name.tsv
done