Merge pull request #2 from erinyoung/erin-dev

working

README.md

@@ -1,8 +1,57 @@
 # roundabout
-Finding regions of similarity between plasmids
 
+The fundamental goal of this repo was to create something quick and reproducible to find large regions of similarity between highly related plasmids - like those in an outbreak.
+
+roundabout uses blast to find regions of similarity (as opposed to groups of genes that would be required for synteny), and then assigns those regions a color. Those regions and colors can then be visualized via circos.
 
 dependencies:
-- blast
-- bedools
-- awk
+- blast : find similarities
+- circos : visualizing the end product (optional)
+- bedtools : combining regions of interest
+- awk/sed/bash : a lot of file manipulation
+- samtools : to find genome size
+
+INSTALL:
+
+```
+git clone https://github.com/erinyoung/roundabout.git
+
+export PATH=$PATH:$(pwd)/roundabout/bin
+
+# Using conda to install dependencies
+conda create -n roundabout -c bioconda -c defaults samtools bedtools circos blast
+
+# then activate the environment with
+conda activate roundabout
+```
+
+USAGE:
+
+- Put the completed/closed plasmids into a single directory (i.e. plasmids)
+- (optional) Put AMRFinderPlus output in a directory (i.e. amrfinder)
+- (optional) Put gff file from prokka/bakta/etc in a directory (i.e. gff)
+
+Note : All files must have the same prefix. `plasmids/$sample.{fasta,fa,fna}`, `amrfinder/$sample*`, `gff/$sample*gff` respectively.
+
+```
+roundabout -d <directory with plasmid sequences>
+
+# example
+roundabout -d plasmids
+
+# with amrfinder results
+roundabout -a amrfinder
+
+# with gff files
+roundabout -g gff
+
+# with both amrfinder results and gff files
+roundabout -a amrfinder -g gff
+```
+
+This was created for personal use with specific projects in mind, as opposed to general use. As such, other users may notice that the use case is highly specific and informal. Put in an issue if this is something that interests you and you need an additional feature.
+
+Future directions:
+- removing the samtools dependency
+- adding parallelization through either gnu parallel or snakemake
+- creating a docker container

bin/roundabout

@@ -12,13 +12,16 @@ Usage:
 roundabout -d <directory of plasmids>
 """
 
-#// TODO : convert to snakefile for parallel processing
-
 roundaboutpath=$(which $0 | sed 's/bin\/roundabout//g')
 
 threads=4
 out="./"
-while getopts "hvd:t:" opt; do
+conf=$roundaboutpath/conf/template.conf
+circos=""
+gff_directory=""
+amrfinder_directory=""
+while getopts "hvd:t:g:a:" opt
+do
   case ${opt} in
     h )
       echo "$USAGE"
@@ -29,22 +32,28 @@ while getopts "hvd:t:" opt; do
       bedtools --version
       samtools --version
       circos --version
-      circlator --version
-      echo "NCBI's AMRFINDERPlus version : " amrfinder --version
       echo "roundabout $VERSION"
       exit 0
       ;;
     d )
       directory=$OPTARG
       if [ ! -d "$directory" ] ; then echo "FATAL : directory with fastas does not exist. Set with '-d'" ; fi
       ;;
+    g )
+      gff_directory=$OPTARG
+      if [ ! -d "$gff_directory" ] ; then echo "FATAL : directory with gff files does not exist. Set with '-g'" ; fi
+      ;;
+    a )
+      amrfinder_directory=$OPTARG
+      if [ ! -d "$amrfinder_directory" ] ; then echo "FATAL : directory with amrfinder results. Set with '-a'" ; fi
+      ;;
     t )
       threads=$OPTARG
       ;;
     o )
       out=$OPTARG
-      mkdir -p $out
-      if [ -d "$out" ] ; then echo "FATAL : Could not create directory for results" ; fi
+      mkdir -p $out/roundabout
+      if [ -d "$out/roundabout" ] ; then echo "FATAL : Could not create directory for results" ; fi
       ;;
     \? )
       echo "$USAGE"
@@ -54,75 +63,86 @@ while getopts "hvd:t:" opt; do
 done
 shift $((OPTIND -1))
 
-if [ -z "$(which blastn)" ]    ; then echo "FATAL : blastn was not found"               ; exit 1 ; fi
-if [ -z "$(which bedtools)" ]  ; then echo "FATAL : bedtools was not found"             ; exit 1 ; fi
-if [ -z "$(which samtools)" ]  ; then echo "FATAL : samtools was not found"             ; exit 1 ; fi
-if [ -z "$(which circlator)" ] ; then echo "FATAL : circlator was not found"            ; exit 1 ; fi
-if [ -z "$(which amrfinder)" ] ; then echo "FATAL : NCBI's AMRFinderPlus was not found" ; exit 1 ; fi
+if [ -z "$(which blastn)" ]   ; then echo "FATAL : blastn was not found"   ; exit 1       ; fi
+if [ -z "$(which bedtools)" ] ; then echo "FATAL : bedtools was not found" ; exit 1       ; fi
+if [ -z "$(which samtools)" ] ; then echo "FATAL : samtools was not found" ; exit 1       ; fi
+if [ -z "$(which circos)" ]   ; then echo "WARNING : circos was not found" ; circos="not" ; fi
 
 mkdir -p $out/roundabout/blast_results
 mkdir -p $out/roundabout/beds
 mkdir -p $out/roundabout/circos
-mkdir -p $out/roundabout/circlator
-mkdir -p $out/roundabout/amrfinder
 
-echo "$(date) : Getting fastas ready"
+if [ -n "$amrfinder_directory" ] ; then conf=$roundaboutpath/conf/template_amr.conf ; fi
+echo "$(date) : conf file is $conf"
 
+echo "$(date) : getting fastas ready"
 prefastas=$(ls $directory/*{.fasta,.fa,.fna} 2> /dev/null)
 for fasta in ${prefastas[@]}
 do
   name=$(basename $fasta | sed 's/.f.*//g')
-  echo "$(date) : rotating $fasta with circlator"
-  circlator fixstart $fasta $out/roundabout/circlator/$name.fixed
-  fold -w 75 $out/roundabout/circlator/$name.fixed.fasta | sed "s/>/>${name}_/g" > $out/roundabout/blast_results/$name.fasta
+  echo "$(date) : formating $fasta"
+  fold -w 75 $fasta | sed "s/>/>${name}_/g" > $out/roundabout/blast_results/$name.fasta
 
   echo "$(date) : getting size of $fasta"
   samtools faidx $out/roundabout/blast_results/$name.fasta
-  cut -f 1,2 $out/roundabout/blast_results/$name.fasta.fai | awk '{print "chr - " $1 " " $1 " 1 " $2 " black"}' > $out/roundabout/circos/${name}_karyotype.txt
+  cut -f 1,2 $out/roundabout/blast_results/$name.fasta.fai | awk '{print "chr - " $1 " " $1 " 0 " $2 " black"}' > $out/roundabout/circos/${name}_karyotype.txt
+
+  echo "$(date) : getting skew for $fasta"
   chr_lengths=$(awk '{print $1 ":" $2 }' $out/roundabout/blast_results/$name.fasta.fai)
   echo -e "#chr\tstr\tend" > $out/roundabout/beds/$name.windows.bed
   for chr_length in ${chr_lengths[@]}
   do
     chr=$(echo $chr_length | cut -f 1 -d ":" )
     length=$(echo $chr_length | cut -f 2 -d ":" )
-    for ((i=1;i<=length;i+=1000))
+    for ((i=1;i<=length;i+=500))
     do
-      if [ "$i" -lt "$((length - 999 ))" ]
+      if [ "$i" -lt "$((length - 499 ))" ]
       then
-        echo -e "$chr\t$i\t$((i + 999 ))" >> $out/roundabout/beds/$name.windows.bed
+        echo -e "$chr\t$i\t$((i + 499 ))" >> $out/roundabout/beds/$name.windows.bed
       else
         echo -e "$chr\t$i\t$length" >> $out/roundabout/beds/$name.windows.bed
       fi
     done
   done
   bedtools nuc -fi $out/roundabout/blast_results/$name.fasta -bed $out/roundabout/beds/$name.windows.bed > $out/roundabout/beds/$name.GC.bed
 
-  echo "$(date) : running $fasta through amrfinder"
-  amrfinder \
-    --nucleotide $out/roundabout/blast_results/$name.fasta \
-    --threads $threads \
-    --name $name \
-    --output $out/roundabout/amrfinder/${name}_amrfinder_plus.txt \
-    --plus
-  cat $out/roundabout/amrfinder/${name}_amrfinder_plus.txt | cut -f 3,4,5,7 > $out/roundabout/beds/${name}_amrfinder.bed
-
-  prokka \
-    --cpu $threads \
-    --outdir $out/prokka/ \
-    --prefix $name \
-    --compliant \
-    $out/roundabout/blast_results/$name.fasta \
-    --force
+  if [ -d "$gff_directory" ]
+  then
+    echo "$(date) : adding bands to karytype file from gff file"
+    cut -f 1,2 $out/roundabout/blast_results/$name.fasta.fai > $out/roundabout/beds/${name}_genome.txt
+    awk '{print $1 "\t" $2 "\t" $2 + 1000000 }' $out/roundabout/beds/${name}_genome.txt > $out/roundabout/beds/$name.genome_subtract.bed
+
+    grep -h ID $gff_directory/${name}*gf* 2> /dev/null | grep -v "region" | awk -v name=$name '{print name "_" $1 "\t" $4 "\t" $5 "\t" $3 }' | sort -k 1,1 -k 2,2n -k 3,3n > $out/roundabout/beds/${name}_bands.bed
+    bedtools merge -i $out/roundabout/beds/${name}_bands.bed | sort -k 1,1 -k 2,2n -k 3,3n > $out/roundabout/beds/${name}_bands_merged.bed
+    bedtools subtract -a $out/roundabout/beds/${name}_bands_merged.bed -b $out/roundabout/beds/$name.genome_subtract.bed > $out/roundabout/beds/${name}_bands_sorted.bed
+
+    grep -e "0$" -e "5$" $out/roundabout/beds/${name}_bands_sorted.bed | awk '{print "band " $1 " " NR " " NR " " $2 " " $3 " gpos25"  }' >  $out/roundabout/beds/${name}_karyotype.bed
+    grep -e "1$" -e "6$" $out/roundabout/beds/${name}_bands_sorted.bed | awk '{print "band " $1 " " NR " " NR " " $2 " " $3 " gpos50"  }' >> $out/roundabout/beds/${name}_karyotype.bed
+    grep -e "2$" -e "7$" $out/roundabout/beds/${name}_bands_sorted.bed | awk '{print "band " $1 " " NR " " NR " " $2 " " $3 " gpos75"  }' >> $out/roundabout/beds/${name}_karyotype.bed
+    grep -e "3$" -e "8$" $out/roundabout/beds/${name}_bands_sorted.bed | awk '{print "band " $1 " " NR " " NR " " $2 " " $3 " gpos100" }' >> $out/roundabout/beds/${name}_karyotype.bed
+    grep -e "4$" -e "9$" $out/roundabout/beds/${name}_bands_sorted.bed | awk '{print "band " $1 " " NR " " NR " " $2 " " $3 " white" }' >> $out/roundabout/beds/${name}_karyotype.bed
+
+    bedtools complement -i $out/roundabout/beds/${name}_bands_merged.bed -g $out/roundabout/beds/${name}_genome.txt > $out/roundabout/beds/${name}_bands_complement.bed
+    bedtools subtract -a $out/roundabout/beds/${name}_bands_complement.bed -b $out/roundabout/beds/$name.genome_subtract.bed > $out/roundabout/beds/${name}_complement_sorted.bed
+    awk '{print "band " $1 " " NR " " NR " " $2 " " $3 " gneg" }' $out/roundabout/beds/${name}_complement_sorted.bed >> $out/roundabout/beds/${name}_karyotype.bed
+
+    sort -k 2,2 -k 5,5n -k 6,6n $out/roundabout/beds/${name}_karyotype.bed | uniq >> $out/roundabout/circos/${name}_karyotype.txt
+  fi
+
+  if [ -d "$amrfinder_directory" ]
+  then
+    echo "$(date) : getting AMR genes $fasta"
+    cat $amrfinder_directory/${name}* 2> /dev/null | awk -v name=$name '{print name "_" $3 " " $4 " " $5 " " $7 }' > $out/roundabout/beds/${name}_amrfinder.bed
+  fi
 done
 
-echo "$(date) : combining NCBI's AMRFinderPlus into one bedfile : $out/roundabout/beds/amrfinder.bed"
-cat $out/roundabout/beds/*_amrfinder.bed > $out/roundabout/beds/amrfinder.bed
-
 echo "$(date) : calculating skew : $out/roundabout/beds/skew.bed"
-grep -v "#" $out/roundabout/beds/*.GC.bed | awk '{$1 " " $2 " " $3 " " $4 }' > $out/roundabout/beds/AC.per.bed
-grep -v "#" $out/roundabout/beds/*.GC.bed | awk '{$1 " " $2 " " $3 " " $5 }' > $out/roundabout/beds/GC.per.bed
+grep -hv "#" $out/roundabout/beds/*.GC.bed | awk '{print $1 " " $2 " " $3 " " $4 }' > $out/roundabout/beds/AC.per.bed
+grep -hv "#" $out/roundabout/beds/*.GC.bed | awk '{print $1 " " $2 " " $3 " " $5 }' > $out/roundabout/beds/GC.per.bed
 #GC Skew is calculated as (G - C) / (G + C)
-grep -v "#" $out/roundabout/beds/*.GC.bed | awk '{$1 " " $2 " " $3 " " $8 - $7 " " $7 + $8 }' | awk '{$1 " " $2 " " $3 " " $4 / $5 }' > $out/roundabout/beds/skew.bed
+grep -hv "#" $out/roundabout/beds/*.GC.bed | awk '{print $1 " " $2 " " $3 " " $8 - $7 " " $7 + $8 }' | awk '{print $1 " " $2 " " $3 " " $4 / $5 }' > $out/roundabout/beds/skew.bed
+
+cat $out/roundabout/beds/*_amrfinder.bed 2> /dev/null | sed "s/-//g" | sed "s/-//g" | sed "s/(//g" | sed "s/)//g" > $out/roundabout/beds/amrfinder.bed
 
 # Getting all of the blast results
 for fasta in $out/roundabout/blast_results/*.fasta
@@ -201,26 +221,27 @@ awk '{print $1 "\t" $2 "\t" $3 "\t" $4 "\t" $5 "\t" $6 "\t" $7 "\t" $8 "\t" $9 "
 bedtools intersect -a $out/roundabout/beds/color_groups.bed -b $out/roundabout/beds/*intersect.bed -f 1 -F 1 -wo > $out/roundabout/beds/blast_color.bed
 
 echo "$(date) : creating final highlights file : $out/roundabout/circos/highlights.bed"
-awk '{if ($13 > $14) print $12 "\t" $13-$2+$10 "\t" $14-$3+$11 "\t" $4 ; if ($13 < $14) print $12 "\t" $13+$2-$10 "\t" $14+$3-$11 "\t" $4 }' $out/roundabout/beds/blast_color.bed | sort -k 1,1 -k 2,2n -k 3,3n | uniq > $out/roundabout/beds/highlights.bed
-cat $out/roundabout/beds/color_groups.bed >> $out/roundabout/circos/highlights.bed
+awk '{if ($13 > $14) print $12 " " $13-$2+$10 " " $14-$3+$11 " " $4 ; if ($13 < $14) print $12 " " $13+$2-$10 " " $14+$3-$11 " " $4 }' $out/roundabout/beds/blast_color.bed | sort -k 1,1 -k 2,2n -k 3,3n | uniq > $out/roundabout/circos/highlights.txt
+cat $out/roundabout/beds/color_groups.bed >> $out/roundabout/circos/highlights.txt
 
 # currently losing samples in highlights.bed, but will need to fix this at the lab
-# additionally, need to look up how to add text and amrfinder genes
-
 echo "$(date) : creating highlighted circos plots for each fasta"
 for karyotype in $out/roundabout/circos/*_karyotype.txt
 do
   name=$(basename $karyotype | sed 's/_karyotype.txt//g')
-  cat $roundaboutpath/conf/template.conf | \
+  cat $conf | \
     sed "s~ROUNDABOUTPATH~${roundaboutpath}/conf~g"      | \
-    sed "s~HIGHLIGHTFILE~${out}/circos/highlights.bed~g" | \
+    sed "s~HIGHLIGHTFILE~$out/roundabout/circos/highlights.txt~g" | \
     sed "s~KARYOTYPEFILE~${karyotype}~g"                 | \
-    sed "s~FINALPNG~${out}/circos/${name}.single.png~g"  | \
+    sed "s~FINALPNG~${out}/roundabout/circos/${name}.single.png~g"  | \
     sed "s~ACSKEWFILE~$out/roundabout/beds/AC.per.bed~g" | \
     sed "s~GCSKEWFILE~$out/roundabout/beds/GC.per.bed~g" | \
     sed "s~SKEWSKEWFILE~$out/roundabout/beds/skew.bed~g" | \
     sed "s~AMRFINDERFILE~$out/roundabout/beds/amrfinder.bed~g" > $out/roundabout/circos/${name}_basic.conf
-  circos -conf $out/roundabout/circos/${name}_basic.conf
+  if [ -z "$circos" ]
+  then
+    circos -conf $out/roundabout/circos/${name}_basic.conf
+  fi
 done
 
 echo "$(date) : roundabout complete!"