generateReport.sh

#! /bin/bash -e

# Prepare an html report. The html file header and standard page banner are copied from a stored file for simplicity.


sampleName=$1
headerFile=$2
isPairedEnd=$3
primerBedFile=$4
projectDir=$5

reportFile=report.html
cp $headerFile $reportFile


IFS='.' read -ra username <<< "$USER"
fullName="${username[0]} ${username[1]}"
timestamp=$(date "+%Y-%m-%d, %T %Z")

echo "<table>" >> $reportFile
echo "<tr>" >> $reportFile
echo "    <td>Sample name:</td><td>$sampleName</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Date generated:</td><td>$timestamp</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Timestamp of C-WAP version used:</td><td>$(git --git-dir $projectDir/.git log -n 1 | grep Date: | sed 's/Date:   //g')</td>" >> $reportFile
echo "</tr>" >> $reportFile


echo "<tr>" >> $reportFile
echo "    <td>Executed by:</td><td>$fullName (<a href=\\"mailto:$USER@fda.hhs.gov?subject=Wastewater report generated on $timestamp\\">$USER@fda.hhs.gov</a>)</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Executed on:</td><td>$(hostname -I | awk '{print $1}') (aka $(hostname))</td>" >> $reportFile
echo "</tr>" >> $reportFile
echo "</table>" >> $reportFile


#######################################################
echo Compiling read statistics
numAligned=$(cat sorted.stats | grep "reads mapped:" | awk '{ print $NF }')
avgQuality=$(cat sorted.stats | grep "average quality" | awk '{ print $NF }')
avgLength=$(cat sorted.stats | grep "average length" | awk '{ print $NF }')

numPassedQuality=$(cat resorted.stats | grep "raw total sequences" | awk '{ print $4 }')
avgQualityPassed=$(cat resorted.stats | grep "average quality" | awk '{ print $NF }')
avgLengthPassed=$(cat resorted.stats | grep "average length" | awk '{ print $NF }')
let "avgCoveragePassed = $avgLengthPassed * $numPassedQuality / 29903" || true


# Deduce the total number of reads from the kraken2 output
# Use taxid column, i.e. column 5
numUnclassified=$(head k2-std.out | awk '$5==0' | awk '{ print $2 }')
if [[ -z $numUnclassified ]]; then
	numUnclassified=0
fi

numClassified=$(head k2-std.out | awk '$5==1' | awk '{ print $2 }')
if [[ -z $numClassified ]]; then
	numClassified=0
fi

# Calculation of total number of reads
let "numReads = $numUnclassified + $numClassified" || true
if $isPairedEnd; then
# Correction for undercounting by kraken2 in the paired mode
	let "numReads = 2 * $numReads" || true
fi


####################################################################
echo >> $reportFile
echo "<br>" >> $reportFile
echo "<h2>Sequencing summary</h2>" >> $reportFile
echo '<table>' >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Sequencing chemistry:</td><td>$libraryProtocol with $seqInstrument</td>" >> $reportFile
echo "</tr>" >> $reportFile

# Catchment site the sample was collected from
echo "<tr>" >> $reportFile
echo "    <td>Source site:</td><td><a href=\"https://www.openstreetmap.org/#map=3/$sampleLatitude/$sampleLongitude\">$sampleLocation\
		($sampleLatitude,$sampleLongitude)</a></td>" >> $reportFile
echo "</tr>" >> $reportFile	

# Date the WW sample was collected from the sanitary system
echo "<tr>" >> $reportFile
echo "    <td>Sampling date:</td><td>$collectionDate</a></td>" >> $reportFile
echo "</tr>" >> $reportFile

# Agency collecting the WW sample
echo "<tr>" >> $reportFile
echo "    <td>Collected by:</td><td>$collectedBy</td>" >> $reportFile
echo "</tr>" >> $reportFile

# Lab that sequenced the sample
echo "<tr>" >> $reportFile
echo "    <td>Sequenced by:</td><td>$sequencedBy</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Total number of reads:</td><td>$numReads</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Reads aligned:</td><td>$numAligned ($(expr 100 \* $numAligned / $numReads)%)</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Average read quality:</td><td>$avgQuality</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Average read length:</td><td>$avgLength</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Reads passing filter:</td><td>$numPassedQuality ($(expr 100 \* $numPassedQuality / $numReads)%)</td>" \
		>> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Average read quality passing filter:</td><td>$avgQualityPassed</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Average read length passing filter:</td><td>$avgLengthPassed</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Average coverage passing filter:</td><td>${avgCoveragePassed}X</td>" >> $reportFile
echo "</tr>" >> $reportFile
echo "</table>" >> $reportFile

echo "<br>" >> $reportFile
echo "A read passes filter if the read length after adaptor trimming &#8805; 30 and minimum \
		read quality &#8805; 20 within a sliding window of width 4." >> $reportFile


#######################################################
echo >> $reportFile
echo "<br><br>" >> $reportFile
echo "<h2>Overall sequence characteristics</h2>" >> $reportFile

echo "<img src=\"./coverage.png\" alt=\"Coverage vs. genome coordinate plot\" width=\"49%\" class=\"center\">" >> $reportFile
echo "<img src=\"./quality.png\" alt=\"Quality vs. genome coordinate plot\" width=\"49%\" class=\"center\">" >> $reportFile
echo "<p> NOTE: The red shaded areas marked with a (<font color="red">*</font>) are not covered by the design of the library preparation kit and hence excluded from analyses. <font color="magenta">Magenta</font> curves represent moving average with a window width of 1kb." >> $reportFile
echo "<br><br><br><br>" >> $reportFile
echo "<img src=\"./depthHistogram.png\" alt=\"Coverage histogram\" width=\"49%\" class=\"center\">" >> $reportFile
echo "<img src=\"./qualityHistogram.png\" alt=\"Quality histogram\" width=\"49%\" class=\"center\">" >> $reportFile
echo "<br><br><br><br>" >> $reportFile
echo "<img src=\"./readLengthHist.png\" alt=\"Read length histogram\" width=\"49%\" class=\"center\">" >> $reportFile
echo "<img src=\"./breadthVSdepth.png\" alt=\"Depth vs breadth plot\" width=\"49%\" class=\"center\">" >> $reportFile
echo "<img src=\"./discontinuitySignal.png\" alt=\"Coverage discontinuity signal\" width=\"49%\" class=\"center\">" >> $reportFile

if [[ -s timeVSreadcounts.png ]]; then 
    echo "<img src=\"./timeVSreadcounts.png\" alt=\"Passage time histogram of ONT reads\" width=\"49%\" class=\"center\">" >> $reportFile
else
    rm timeVSreadcounts.png
fi

echo >> $reportFile
echo "<br>" >> $reportFile

# Check for coverage data. If the coverage is too low, any variant calling will be inaccurate
if [ $avgCoveragePassed -lt 10 ]; then
	echo "<p> <font color=\"red\"> WARNING: The sequence coverage is very low (${avgCoveragePassed}X) </font>" >> $reportFile
fi



#######################################################
# Check the coverage depth across reference sequence
# Multiple metrics are reported

# Evaluation of number of genomic coordinates that could not be amplified due to being
# out of range of the amplicon kit.
echo Assessing completeness of coverage...
numUncoveredLociByDesign=0
while read line; do
    gapBegin=$(echo $line | awk '{print $1}')
    gapEnd=$(echo $line | awk '{print $2}')
    let "numUncoveredLociByDesign += $gapEnd - $gapBegin + 1"
done <<< "$(python3 $projectDir/findUncoveredCoordinates.py $primerBedFile 1)"


poorlyCoveredLoci=($(cat pos-coverage-quality.tsv | awk '$2 < 10 { print $1 }'))
uncoveredLoci=($(cat pos-coverage-quality.tsv | awk '$2 == 0 { print $1 }'))

# Coverage analysis for SNP loci currently in circulation
# Substitutions present >=90% of GISAID submissions
commonSNPuncov=()
commonSNPpoorlycov=()
numCommonSNP=$(cat $projectDir/recentVariants/commonSNPpos | wc -l)
for locus in $(cat $projectDir/recentVariants/commonSNPpos); do
	if echo ${poorlyCoveredLoci[@]} | grep -q -w $locus; then
		commonSNPpoorlycov+=(locus)
	fi
	if echo ${uncoveredLoci[@]} | grep -q -w $locus; then
		commonSNPuncov+=(locus)
	fi
done

# Substitutions present (10%,90%) of GISAID submissions
rareSNPuncov=()
rareSNPpoorlycov=()
numDiverseSNP=$(cat $projectDir/recentVariants/diverseSNPpos | wc -l)
for locus in $(cat $projectDir/recentVariants/diverseSNPpos); do
	if echo ${poorlyCoveredLoci[@]} | grep -q -w $locus; then
		diverseSNPpoorlycov+=(locus)
	fi
	if echo ${uncoveredLoci[@]} | grep -q -w $locus; then
		diverseSNPuncov+=(locus)
	fi
done

# Substitutions present <=10% of GISAID submissions
rareSNPuncov=()
rareSNPpoorlycov=()
numRareSNP=$(cat $projectDir/recentVariants/rareSNPpos | wc -l)
for locus in $(cat $projectDir/recentVariants/rareSNPpos); do
	if echo ${poorlyCoveredLoci[@]} | grep -q -w $locus; then
		rareSNPpoorlycov+=(locus)
	fi
	if echo ${uncoveredLoci[@]} | grep -q -w $locus; then
		rareSNPuncov+=(locus)
	fi
done

echo '<table>' >> $reportFile
echo "<tr>" >> $reportFile
echo "    <th></th><th>Uncovered coordinates (0X)</th><th>Poorly covered coordinates (<10X)</th>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td># Inaccessible genomic coordinates by kit design:</td><td>${numUncoveredLociByDesign}nt ($(expr 100 \* $numUncoveredLociByDesign / 29903)%)</td><td>${numUncoveredLociByDesign}nt ($(expr 100 \* $numUncoveredLociByDesign / 29903)%)</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>All genomic coordinates:</td><td>${#uncoveredLoci[@]}nt ($(expr 100 \* ${#uncoveredLoci[@]} / 29903)%)</td><td>${#poorlyCoveredLoci[@]}nt ($(expr 100 \* ${#poorlyCoveredLoci[@]} / 29903)%)</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Common SNPs:</td><td>${#commonSNPuncov[@]}nt ($(expr 100 \* ${#commonSNPuncov[@]} / $numCommonSNP)%)</td>" \
				"<td>${#commonSNPpoorlycov[@]}nt ($(expr 100 \* ${#commonSNPpoorlycov[@]} / $numCommonSNP)%)</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Diverse SNPs:</td><td>${#diverseSNPuncov[@]}nt ($(expr 100 \* ${#diverseSNPuncov[@]} / $numDiverseSNP)%)</td>" \
				"<td>${#diverseSNPpoorlycov[@]}nt ($(expr 100 \* ${#diverseSNPpoorlycov[@]} / $numDiverseSNP)%)</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile	
echo "    <td>Rare SNPs:</td><td>${#rareSNPuncov[@]}nt ($(expr 100 \* ${#rareSNPuncov[@]} / $numRareSNP)%)</td>" \
				"<td>${#rareSNPpoorlycov[@]}nt ($(expr 100 \* ${#rareSNPpoorlycov[@]} / $numRareSNP)%)</td>" >> $reportFile
echo "</tr>" >> $reportFile
echo '</table>' >> $reportFile

echo "SNPs refer to the polymorphic sites currently in circulation that were detected out of recent GISAID entries. The sites that differ from the SC2 reference sequence are denoted as \"common\" if [90%, 100%] of the submissions carry this mutation, whereas those that are prevalent in [0%,10%] of the submissions are grouped under the \"rare\" category. The population is still diverse at the mutation sites that are observed in (10%,90%) of the entries and these coordinates are grouped under the \"diverse\" category." >> $reportFile


# Plots of missing spot counts, grouped w.r.t. SC2 genes.
echo "<br><br><br><br>" >> $reportFile
echo "<img src=\"./genesVSuncovered_abscounts.png\" alt=\"Uncovered genome coordinates per gene, absolute counts\" width=\"49%\" class=\"center\">" >> $reportFile
echo "<img src=\"./genesVSuncovered_scaled.png\" alt=\"Uncovered genome coordinates per gene, counts per kb\" width=\"49%\" class=\"center\">" >> $reportFile
echo "<br><br><br><br>" >> $reportFile



#######################################################
# Taxonomic classification results of the reads
numCovid=$(cat k2-std.out | grep Orthocoronavirinae | grep -v unclass | awk '{ print $2 }')
numHuman=$(cat k2-std.out | grep sapiens | awk '{ print $2 }')
numSynthetic=$(cat k2-std.out | grep other | grep 28384 | awk '{ print $2 }')

if [[ -z $numCovid ]]; then
	pctCovid=0.00
	numCovid=0
else
	pctCovid=$(cat k2-std.out | grep Orthocoronavirinae | grep -v unclass | awk '{ print $1 }')
fi

if [[ -z $numHuman ]]; then
	pctHuman=0.00
	numHuman=0
else
	pctHuman=$(cat k2-std.out | grep sapiens | awk '{ print $1 }')
fi

if [[ -z $numSynthetic ]]; then
	pctSynthetic=0.00
	numSynthetic=0
else
	pctSynthetic=$(cat k2-std.out | grep other | grep 28384 | awk '{ print $1 }')
fi

k2hits=$(cat k2-std.out | sort -k 1 -rn | awk '$4=="F" {print $6 " (" $1 "%)<br>"}' | head -n 5)

echo '<table>' >> $reportFile
echo "<tr>" >> $reportFile
echo "    <td>Hits to SARS-Cov2 genome (kraken2):</td><td>$numCovid reads (${pctCovid}%) </td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Hits to human genome (kraken2):</td><td>$numHuman reads (${pctHuman}%)</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Hits to synthetic sequences (kraken2, taxid 28384):</td><td>$numSynthetic reads (${pctSynthetic}%)</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "<tr>" >> $reportFile
echo "    <td>Most abundant organisms (kraken2, family level):</td><td>$k2hits</td>" >> $reportFile
echo "</tr>" >> $reportFile

echo "</table>" >> $reportFile



#########################################################
# QC classification
# VVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVV
qc_flags=./qc-flags.txt
touch $qc_flags

dominantOrganism=$(echo $k2hits | head -n 1 | awk '{print $1}')
if [[ $dominantOrganism != "Coronaviridae" ]]; then
	echo 'unspecific_amplification' >> $qc_flags
fi

if [[ $numReads -le 100 ]]; then
	echo 'low_sequencing_depth' >> $qc_flags
fi

# Avg. coverage < 2X -> error, F
# Avg. coverage < 1000X -> warning, B
# Avg. coverage < 100X -> warning, C
if [[ $avgCoveragePassed -lt 2 ]]; then
	echo 'insufficient_average_coverage' >> $qc_flags
else
	if [[ $avgCoveragePassed -lt 100 ]]; then
		echo 'low_average_coverage' >> $qc_flags
	else
		if [[ $avgCoveragePassed -lt 1000 ]]; then
			echo 'suboptimal_average_coverage' >> $qc_flags
		fi
	fi
fi

if [[ $avgLengthPassed -le 70 ]]; then
	echo 'truncated_reads' >> $qc_flags
fi

# 0x coordinates > 1%
if [[ ${#uncoveredLoci[@]} -gt 300 ]]; then
	echo 'low_coverage_breadth' >> $qc_flags
fi

# 10X coordinates > 5%
if [[ ${#uncoveredLoci[@]} -gt 1500 ]]; then
	echo 'uneven_sequence_coverage' >> $qc_flags
fi

# Adjust for PB/Nanopore
if [[ $avgQualityPassed -lt 30 ]]; then
	echo 'sequence_quality_low' >> $qc_flags
fi

# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^



#######################################################
echo "<br>" >> $reportFile
echo "<h2>Detected variants (Experimental)</h2>" >> $reportFile
echo "<div>" >> $reportFile
echo "    <div id=\"figdiv\">" >> $reportFile
echo "        <img src=\"./pieChart_deconvolution.png\" alt=\"Abundance of variants by deconvolution\" width=\"100%\">" >> $reportFile
echo "    </div>" >> $reportFile
echo "    <div id=\"figdiv\">" >> $reportFile
echo "        <img src=\"./pieChart_kallisto.png\" alt=\"Abundance of variants by kallisto\" width=\"100%\">" >> $reportFile
echo "    </div>" >> $reportFile
echo "</div>" >> $reportFile

echo "<div id=\"caption\">" >> $reportFile
echo "    <p> Based on deconvolution, <a href=\"https://outbreak.info/situation-reports?pango=$mostAbundantVariantName\">$mostAbundantVariantName</a> 
	is estimated to constitute $mostAbundantVariantPct% of the viral\
	particles and hence is the most abundant variant in the sample. The R<sup>2</sup> for the linear regression was $linRegressionR2.\
	Variants that were detected less than 5% were grouped under \"Other\"" >> $reportFile
echo "    <p> Based on the consensus sequence of the observed reads, the \"ensemble-averaged sequence\" most closely resembles \
	the <a href=\"https://outbreak.info/situation-reports?pango=$consensusLineage\">$consensusLineage</a> lineage. \
	If this is a sample consisting of a single source of pathogens or an overwhelming majority of the different sources \
	are infected with the same variant, the sample is dominated by this variant." >> $reportFile
echo "    <p> Based on mapping individual reads to the variant consensus sequences in the reference database, \
	kallisto predicts that the sample is dominated by <a href=\"https://outbreak.info/situation-reports?pango=$kallistoTopName\"\
	>$kallistoTopName</a> lineage. Accuracy of this measure is expected to improve if the \
	input data consists of long reads as opposed to convolution." >> $reportFile
echo "</div>" >> $reportFile
echo "<br>" >> $reportFile
	
echo "<br>" >> $reportFile
echo "<div>" >> $reportFile
echo "    <div id=\"figdiv\">" >> $reportFile
echo "        <img src=\"./pieChart_k2_allCovid.png\" alt=\"Abundance of variants by kraken2+bracken\"\
					width=\"100%\">" >> $reportFile
echo "    </div>" >> $reportFile
echo "    <div id=\"figdiv\">" >> $reportFile
echo "        <img src=\"./pieChart_k2_majorCovid.png\" alt=\"Abundance of variants by kraken2+bracken\"\
					width=\"100%\">" >> $reportFile
echo "    </div>" >> $reportFile
echo "</div>" >> $reportFile
echo "<br>" >> $reportFile

echo "<br>" >> $reportFile
echo "<div>" >> $reportFile
echo "    <div id=\"figdiv\">" >> $reportFile
echo "        <img src=\"./pieChart_freyja.png\" alt=\"Abundance of variants by Freyja\"\
				width=\"100%\">" >> $reportFile
echo "    </div>" >> $reportFile
echo "    <div id=\"figdiv\">" >> $reportFile
echo "        <img src=\"./pieChart_lcs.png\" alt=\"Abundance of variants by LCS\"\
				width=\"100%\">" >> $reportFile
echo "    </div>" >> $reportFile
echo "</div>" >> $reportFile
echo "<br>" >> $reportFile


echo "<br>" >> $reportFile
echo "<div>" >> $reportFile
echo "    <div id=\"figdiv\">" >> $reportFile
echo "        <img src=\"./freyja_bootstrap.png\" alt=\"Freyja bootstrapping\"\
				width=\"100%\">" >> $reportFile
echo "    </div>" >> $reportFile
echo "    <div id=\"figdiv\">" >> $reportFile
echo "        " >> $reportFile
echo "    </div>" >> $reportFile
echo "</div>" >> $reportFile
echo "<br>" >> $reportFile

# Append the VOC - VOI support table generated by the above Python script to the report
cat VOC-VOIsupportTable.html >> $reportFile
rm VOC-VOIsupportTable.html



#######################################################
echo Appending a detailed list of all detected mutations...
echo >> $reportFile
echo "<h2>Detected mutations</h2>" >> $reportFile
echo "Only genomic coordinates with at least 10X coverage were considered." >> $reportFile
echo "<br>" >> $reportFile

echo "<table>" >> $reportFile
echo "<tr>" >> $reportFile
echo "    <th>Position</th> <th>Ref. base</th> <th>Alt. base</th> <th>Alt. freq</th> <th>p-value</th> <th>Mutation name</th> <th>Compatible lineages</th>" >> $reportFile
echo "</tr>" >> $reportFile
	
# Copy the list of mutations from the temporary file generated by python
cat mutationTable.html >> $reportFile
rm mutationTable.html
echo "</table>" >> $reportFile
echo "<br>" >> $reportFile
	
echo >> $reportFile
echo "</body>" >> $reportFile
echo "</html>" >> $reportFile



# Re-organise the folder so that it is ready to export.
mkdir report
mv report.* report/
mv *.png report/

mkdir outfolder
shopt -s extglob
mv ./!(outfolder) ./outfolder/