path: root/SelectiveGenomeAmplification

#!/usr/bin/env bash


all=run
# Parse in our arguments
if [[ -z "$foreground" ]] && [[ -z "$background" ]]; then 
	if (( $# < 2 )); then  
		echo "please supply two genomes, foreground and background"
		exit 1
	fi;

	: ${foreground=$1}
	: ${background=$2}
	start=3
else
	start=1
fi

if (( $# > 2 )); then
	unset all

	for i in "${@:$start}"; do 
		if [[ "$i" = "1" ]] || [[ "$i" = "count" ]]; then
			step=1;	step_mers=1
		fi

		if [[ "$i" = "2" ]] || [[ "$i" = "filter" ]]; then
			step=1;	step_filters=1
		fi

		if [[ "$i" = "3" ]] || [[ "$i" = "select" ]]; then
			step=1;	step_select=1
		fi

		if [[ "$i" = "4" ]] || [[ "$i" = "score" ]]; then
			step=1; step_score=1
		fi

		if [[ $step ]] && [[ ! "$current_run" ]] && [[ ! $step_mers ]]; then
			echo "Error: If you are going to step through your program, and aren't starting" \
			"at the first step, you better specify what previous run you want to use" \
			"as your base"
			exit
		fi

		done;
fi;

if [[ -n "$step" ]] && [[ -z "$step_mers" ]] &&  [[ -z "$step_filters" ]] && [[ -z $step_select ]] && [[ -z "$step_score" ]]; then
	echo "Error: you need to select at least one step to run."
	exit 
fi

echo 
echo "Planning on running these steps:"
for var in step_mers step_filters step_select step_score all; do
	if [[ -n "${!var}" ]]; then
		echo '  '$var
	fi
done

# output directory 
: ${output_directory=$(basename "$foreground")_$(basename "$background")}

# temp directory 
: ${tmp_directory="$output_directory"/.tmp}

# directory to store our counts and sorted counts
: ${counts_directory="$tmp_directory"}

# range of mers, min and max 
: ${min_mer_range=6}
: ${max_mer_range=12}

# max mer distance, the distance between two mers in our selected outputs
: ${max_mer_distance=5000}

# min/maximum kmer meling point
: ${max_melting_temp=30}
: ${min_melting_temp=0}

# minimum average binding distance in the foreground
: ${min_foreground_binding_average=50000}

# maximum mers to pick
: ${max_select=15}

# maximum mers to check
: ${max_check=35}

# mers to specifically IGNORE, space delimited
: ${ignore_mers=''}

# maximum number of mers that are consecutively binding
: ${max_consecutive_binding=4}

# fg_weight, how much to weight to give the higher bindnig primers
: ${fg_weight=0}

# primer_weight, how much weight to give to sets with a higher number of primers. (between 0 and 1)
: ${primer_weight=0}

export ignore_mers
export min_mer_range
export max_mer_range

export max_select

export min_foreground_binding_average
export max_mer_distance

export max_melting_temp 
export min_melting_temp 

export fg_weight
export primer_weight


echo
# check foreground and background
if [[ ! -f "$foreground" ]]; then
	echo "Error: could not open $foreground"
	exit 1
fi

if [[ ! -f "$background" ]]; then
	echo "Error: could not open $background"
	exit 1
fi


if [[ -n "$current_run" ]] && [[ ! -d "$output_directory/$current_run" ]]; then
	echo -n "run $current_run was not found, it should be a folder here: "
	echo "$output_directory/$current_run"
	exit
fi

num=1
if [[ -z "$current_run" ]]; then 
	while [[ -d $output_directory/run_$num ]] ; do
		let num++
	done
	current_run=run_$num
fi

fg_basename=$(basename "$foreground")
bg_basename=$(basename "$background")

fg_counts=$counts_directory/$fg_basename-counts
bg_counts=$counts_directory/$bg_basename-counts
selected=$output_directory/$current_run/selected-mers

	 
average_binding=$output_directory/$current_run/$fg_basename-counts-average-binding
non_melting=$output_directory/$current_run/$fg_basename-counts-non-melting
consecutive_binding=$output_directory/$current_run/$fg_basename-counts-consecutive-binding

# Make our output directory
if [[ ! -d "$output_directory" ]]; then
	mkdir "$output_directory"
fi

# Make our counts directory
if [[ ! -d "$counts_directory" ]]; then
	mkdir "$counts_directory"
fi

# Make our temporary directory
if [[ ! -d $tmp_directory ]]; then
	mkdir "$tmp_directory"
fi

# Make our current run directory 
if [[ ! -d $output_directory/$current_run ]]; then
	mkdir "$output_directory"/"$current_run" 
fi

echo "Outputting current run parameters"
	for var in ignore_mers min_mer_range max_mer_range max_check cpus max_consecutive_binding max_select min_foreground_binding_average max_mer_distance min_melting_temp max_melting_temp foreground background; do 
	echo "$var" "${!var}" >> "$output_directory"/"$current_run"/parameters
done;

echo "current run is: $current_run"
echo

if [[ -n "$step_mers" ]] || [[ -n "$all" ]]; then
	# to continue this project you need to use the current run.

	echo "Step 1: counting primers in foreground and background"

	for fasta_file in "$foreground" "$background"; do

		counts="$counts_directory"/$(basename "$fasta_file")

		echo "counting mers in $fasta_file"

		# check each mer size and process if not already run
		for (( mer = min_mer_range; mer <= max_mer_range; mer++)) ; do 
			if [[ ! -e "$counts"-counts-"$mer" ]]; then
				echo "checking $mer mers for $fasta_file (assuming $fasta_file didn't change)"
				kmer_continuous_count -c -i "$fasta_file" -k "$mer" -l -n > "$counts"-counts-"$mer" || exit 1
			else 
				echo "$mer mers already done for $fasta_file"
			fi

		#  remove the counts file so we can concatenate
		if [[ -e "$counts"-counts ]]; then
			rm "$counts"-counts
		fi
			
		# concatentate 
		cat "$counts"-counts-"$mer" >> "$counts"-counts
		
		done
	done
fi

if [[ -n "$step_filters" ]] || [[ -n "$all" ]]; then
	if [[ ! -f "$fg_counts" ]]; then
		echo "Error: you need to run your count step before filtration"
		exit
	fi
	echo "Step 2: Filtering mer combinations based on parameters"

	# remove ignored mers
	if [[ "$ignore_mers" ]]; then
		echo "removing ignored mers: " + "$ignore_mers"
		for mer in $ignore_mers; do
			sed -i '/^'"$mer"'\t/d' "$fg_counts"
			sed -i '/^'"$mer"'\t/d' "$bg_counts"
		done
	fi

	echo "checking if mers appear at least as often in the fg as the average binding site or more $min_foreground_binding_average"
	filter_average_binding.py "$foreground" "$min_foreground_binding_average" < "$fg_counts"  > "$average_binding" || exit 1

	echo "checking if mers are within the melting range $min_melting_temp $max_melting_temp"
	filter_melting_temperature.py "$min_melting_temp" "$max_melting_temp" < "$average_binding" > "$non_melting" || exit 1

	echo "filtering out elements that have more consecutive binding mers than allowed by \$max_consecutive_binding $max_consecutive_binding"
	filter_max_consecutive_binding.py "$max_consecutive_binding" < "$non_melting" > "$consecutive_binding" || exit 1
fi

if [[ -n "$step_select" ]] || [[ -n "$all" ]]; then
	if [[ ! -f "$consecutive_binding" ]]; then
		echo "Error: you need to run your filtration step before selection"
		exit
	fi
	echo "Step 3: Scoring mer selectivity"
	select_mers.py "$consecutive_binding" "$bg_counts" > "$selected" || exit 1
fi

if [[ -n "$step_score" ]] || [[ -n "$all" ]]; then
	if [[ ! -f "$selected" ]]; then
		echo "Error: you need to run your selection step before you run your scoring"
		exit
	fi

	echo "Step 4: Scoring top mers based on selectivity"
	score_wrapper.sh "$selected" "$foreground" "$background" "$output_directory"/"$current_run"/scores-output || exit 1
fi