Updated index script to generate whippet index

core-w/2020-10-18_makeIndex.sh

@@ -3,7 +3,7 @@
 # written by A.J. Sethi on 2020-10-18
 
 # testing;
-# module load clairo; bash ~/ClaiRO/core-w/2020-10-18_makeIndex.sh -a "/g/data/lf10/as7425/SRscan/genome_dmel-r6.35" -w "/home/150/as7425/.julia/v0.6/Whippet/bin/" -n "/scratch/lf10/as7425/nascent_bam" -c "/scratch/lf10/as7425/cytoplasmic_bam" -o "/scratch/lf10/as7425/clairo_out"
+# module load clairo; bash ~/ClaiRO/core-w/2020-10-18_makeIndex.sh -a "/g/data/lf10/as7425/genomes/human_genome" -w "/home/150/as7425/.julia/v0.6/Whippet/bin/" -n "/g/data/lf10/as7425/2020-10-20_castilloGuzman_analysis/clairoInput/nascentAlignments" -c "/g/data/lf10/as7425/2020-10-20_castilloGuzman_analysis/clairoInput/cytoplasmicAlignments" -o "/g/data/lf10/as7425/2020-10-20_castilloGuzman_analysis/clairoOutput" -t "48" -v "TRUE"
 
 
 # submit with;
@@ -176,27 +176,58 @@ vsec "completed housekeeping; parsing GTF"
 ####################################
 
 # generate on the fly whippet index using mature bams
+nsec "preparing cytoplasmic alignments for whippet indexing"
 
 # first, merge the mature bams
 matureSamMerge="${od}/matureSamMerge"; mkdir -p ${matureSamMerge} 2>/dev/null
-cd ${cbamDir} && samtools merge -f -@ ${threadCount} ${matureSamMerge}/merged.bam `ls *.bam` 2>/dev/null || die "cannot merge the cytoplasmic RNA alignments"
-samtools sort -@ ${threadCount} ${matureSamMerge}/merged.bam > ${matureSamMerge}/sorted.merged.bam || die "cannot sort your bam"
-samtools rmdup -S ${matureSamMerge}/sorted.merged.bam ${matureSamMerge}/unique.sorted.merged.bam || die "cannot remove duplicates"
-samtools index -@ ${threadCount} ${matureSamMerge}/unique.sorted.merged.bam || die "cannot index your merged bam"
 
+# merge bams
+vsec "preparing to merge mature alignments"
+if [ -f "${matureSamMerge}/merged.bam" ];
+then vsec "previous merged alignment found -- skipping merge step";
+else
+cd ${cbamDir} && samtools merge -f -@ ${threadCount} ${matureSamMerge}/merged.bam `ls *.bam` 2>/dev/null || die "cannot merge the cytoplasmic RNA alignments";
+fi
 
-ssec "fasta is ${myFasta}"
-ssec "bam is "${matureSamMerge}/unique.sorted.merged.bam""
-ssec "annotation is ${myAnnotation}"
+# sort the merged bams
+vsec "preparing to sort your merged mature alignments"
+if [ -f "${matureSamMerge}/sorted.merged.bam" ];
+then vsec "previous merged alignment found -- skipping sort step";
+else
+samtools sort -@ ${threadCount} ${matureSamMerge}/merged.bam > ${matureSamMerge}/sorted.merged.bam || die "cannot sort your bam"
+fi
 
+# remove duplicates from sorted, merged bam
+# samtools is poorly optimized
+# so we first split the bam into chromosomes
+# then remove duplicates
+# and finally remerge everything
+vsec "preparing to removed duplicates from sorted + merged alignments"
+if [ -f "${matureSamMerge}/unique.sorted.merged.bam" ];
+then vsec "previous unique alignment found -- skipping deduplication step";
+else # proceed to split the bam
+
+  samtools rmdup -S ${matureSamMerge}/sorted.merged.bam ${matureSamMerge}/unique.sorted.merged.bam || die "cannot remove duplicates"
+fi
 
+# index the unique bam
+vsec "indexing sunique alignments"
+if [ -f "${matureSamMerge}/unique.sorted.merged.bam.bai" ];
+then vsec "previous index found -- skipping indexing step";
+else samtools index -@ ${threadCount} ${matureSamMerge}/unique.sorted.merged.bam || die "cannot index your merged bam"
+fi
+
+### testing
+### tell me the files we're using for the whippet indexing step
+vsec "notifications incoming:"
+vsec "fasta is ${myFasta}"
+vsec "bam is "${matureSamMerge}/unique.sorted.merged.bam""
+vsec "annotation is ${myAnnotation}"
 
 # ask whippet to make an index while recognizing splice junctions from our bam
 wptIdx="/g/data/lf10/as7425/SRscan/index-whippet"; mkdir ${wptIdx} 2>/dev/null; cd ${wptIdx} || die "cannot access whippet index directory"; rm -rf ${wptIdx}/*
 julia ${whippetPath}/whippet-index.jl --fasta "${myFasta}" --bam "${matureSamMerge}/unique.sorted.merged.bam" --gtf "${myAnnotation}" --bam-min-reads 3  --index ${wptIdx} || die "canot make whippet index"
 
-
-
 exit 0
 ####################################