benchmarks relating to SBT-allsome publication

paper_benchmarks/all_benchmarks.sh

@@ -0,0 +1,196 @@
+#!/bin/bash
+
+special_thousand=0
+special_gencode=0
+
+for dataset in individual_queries/individual 
+#ten hundred
+do
+
+
+if [ $dataset == "ten" ]
+then
+lasti=9
+else
+	if [ $dataset == "individual_queries/individual" ]
+	then
+		lasti=50
+	else
+		lasti=2
+	fi
+fi
+
+
+for i in `seq 0 $lasti`
+do 
+
+if [ $special_thousand == "1" ]
+then
+	query=thousand
+else
+	if [ $special_gencode == "1" ]
+	then
+		query=gencode.v25.transcripts
+	else
+		query=$dataset.$i
+fi
+fi
+
+echo "query: $query"
+
+
+# query prep
+
+cat /mnt/assemblage/sbt/$query.fa \
+  | grep -v "^>" \
+  > temp/$query.query
+
+# original
+
+do_original=1
+
+if [ $do_original == "1" ]
+then
+
+	bash drop_cache > /dev/null
+	echo "original SK"
+
+	/usr/bin/time ../src/bt query-original \
+	  --query-threshold 0.9 \
+	  /mnt/assemblage/sbt/compressedSBT/SBT_list.txt \
+	  temp/$query.query \
+	  dat/$query.unclust-rrr-original.dat \
+	   > log/$query.unclust-original  2>&1
+
+	tail -n 2 log/$query.unclust-original |head -n 1
+fi
+
+# original + opt
+
+do_original_opt=1
+
+if [ $do_original_opt == "1" ]
+then
+
+	bash drop_cache > /dev/null
+	echo "original SK + opt"
+
+	/usr/bin/time ../src/bt query-redux \
+	  --query-threshold 0.9 \
+	  /mnt/assemblage/sbt/compressedSBT/SBT_list.txt \
+	  temp/$query.query \
+	  dat/$query.unclust-rrr-original.dat \
+	   > log/$query.unclust-original-opt  2>&1
+
+	tail -n 2 log/$query.unclust-original-opt |head -n 1
+fi
+
+# clust alone
+
+do_clust_alone=1
+
+if [ $do_clust_alone == "1" ]
+then
+	bash drop_cache > /dev/null
+	echo "clust"
+
+	/usr/bin/time ../src/bt query-original \
+	  --query-threshold 0.9 \
+	  /mnt/assemblage/sbt/clusteringSBT.conversion/sbt-rrr.txt \
+	  temp/$query.query \
+	  dat/$query.clust-rrr-original.dat \
+	  > log/$query.clust-original 2>&1
+
+	tail -n 2 log/$query.clust-original |head -n 1
+
+fi
+
+# old clust+allsome
+
+#echo tableId      = SBT+clust+allsome
+#echo queryId      = clust-rrr-split
+#echo queryProcess = query-redux
+#echo topologyFile = /gpfs/cyberstar/pzm11/backup/sbt/clusteringSBT.compressed/sbt-rrr-split.txt
+
+#/usr/bin/time ../src/bt query-redux \
+#  --query-threshold 0.9 \
+#  /mnt/assemblage/sbt/clusteringSBT.compressed/sbt-rrr-split.txt \
+#  temp/$query.query \
+#  clust-rrr-split.$query.dat > log/$query.split  2>&1
+
+#tail -n 2 log/$query.split | head -n 1
+
+# original SK + allsome
+
+do_original_allsome=1
+
+if [ $do_original_allsome == "1" ]
+then
+
+	bash drop_cache > /dev/null
+	echo "original+allsome"
+
+	/usr/bin/time ../src/bt query-redux \
+	  --query-threshold 0.9 \
+	  /mnt/assemblage/sbt/compressedSBT.allsome/sbt-rrr-allsome.txt \
+	  temp/$query.query \
+	  dat/$query.unclust-rrr-allsome.dat \
+	   > log/$query.unclust-allsome 2>&1
+
+	tail -n 2 log/$query.unclust-allsome | head -n 1
+fi
+
+# original SK + allsplit-split(2 files)
+
+do_original_allsome_split=1
+
+if [ $do_original_allsome_split == "1" ]
+then
+
+	bash drop_cache > /dev/null
+	echo "original+allsome-split"
+
+	/usr/bin/time ../src/bt query-redux \
+	  --query-threshold 0.9 \
+	  /mnt/assemblage/sbt/compressedSBT.reconstructed/sbt-rrr-split.txt \
+	  temp/$query.query \
+	  dat/$query.unclust-rrr-split.dat \
+	   > log/$query.unclust-split 2>&1
+
+	tail -n 2 log/$query.unclust-split | head -n 1
+fi
+
+
+
+# new clust+allsome
+
+do_clust_allsome=1
+
+if [ $do_clust_allsome == "1" ]
+then
+
+	bash drop_cache > /dev/null
+	echo "clust+allsome (1-file)"
+
+	/usr/bin/time ../src/bt query-redux \
+	  --query-threshold 0.9 \
+	  /mnt/assemblage/sbt/clusteringSBT.compressed/sbt-rrr-allsome.txt \
+	  temp/$query.query \
+	  dat/$query.clust-rrr-allsome.dat \
+	   > log/$query.clust-allsome 2>&1
+
+	tail -n 2 log/$query.clust-allsome | head -n 1
+fi
+
+
+
+
+if [ $special_thousand == "1" ] || [ $special_gencode == "1" ]
+then
+exit 0
+fi
+
+done
+done
+
+

paper_benchmarks/check-correctness.sh

@@ -0,0 +1,5 @@
+grep "^*" clust-rrr-original.ten.6.dat |sort > clust-rrr-original.ten.6.sequences.dat
+grep "^*" clust-rrr-split.ten.6.dat |sort > clust-rrr-split.ten.6.sequences.dat
+grep "^*" clust-rrr-allsome.ten.6.dat |sort > clust-rrr-allsome.ten.6.sequences.dat
+diff clust-rrr-original.ten.6.sequences.dat clust-rrr-split.ten.6.sequences.dat
+diff clust-rrr-split.ten.6.sequences.dat clust-rrr-allsome.ten.6.sequences.dat

paper_benchmarks/drop_cache

@@ -0,0 +1 @@
+sudo bash -c "free && sync && echo 3 > /proc/sys/vm/drop_caches && free"

paper_benchmarks/generate_individual.sh

@@ -0,0 +1,10 @@
+#!/bin/sh
+rm individual_queries/*.fa
+seqtk sample gencode.v25.transcripts.fa 50 > individual_queries/50_queries.fa
+cd individual_queries
+split -l 2 50_queries.fa
+num=0
+for file in x*; do
+       mv "$file" "individual.$(printf "%u" $num).fa"
+       let num=$num+1
+done