cds/tools/chlorodb added

Former-commit-id: 0579e878a69b7c285ca71870e9ca5730649a2fda
Former-commit-id: 7cced5b488441d87bf070a9a444317db0e048880

detectors/cds/tools/chlorodb/subdb/go_subdb.sh

@@ -0,0 +1,195 @@
+#!/bin/csh -f
+#
+# usage: go_subdb.sh prot.fst pat.txt [deltalen covmin pmax idmin sizmin]
+# usage: prot.fst : proteins fasta file
+# usage: pat.txt  : text file containing patterns and names for families to extract
+# usage: output directory containig subdbs : basename <pat:r>.db
+#
+
+unsetenv ORG_SOURCED
+
+setenv ORG_HOME `dirname $0`/../../../../..
+source $ORG_HOME/scripts/csh_init.sh
+
+NeedArg 2
+
+set ProtFile = $Argv[1]; Shift
+set PatFile  = $Argv[1]; Shift
+
+NeedFile $ProtFile
+NeedFile $PatFile
+
+#
+# parameters
+#
+
+set Delta  = 0.5
+set Covmin = 30
+set Pmax   = 1e-6
+set Idmin  = 30
+set Sizmin = 5
+
+if ($#Argv > 0) then
+  set Delta = $Argv[1]; Shift
+endif
+
+if ($#Argv > 0) then
+  set Covmin = $Argv[1]; Shift
+endif
+
+if ($#Argv > 0) then
+  set Pmax = $Argv[1]; Shift
+endif
+
+if ($#Argv > 0) then
+  set Idmin = $Argv[1]; Shift
+endif
+
+if ($#Argv > 0) then
+  set Sizmin = $Argv[1]; Shift
+endif
+
+#
+# output directory
+#
+
+set OutDir = `basename $PatFile:r`.db
+
+if (-d $OutDir) \rm -r $OutDir
+mkdir $OutDir
+
+set OutLog = `basename $PatFile:r`.log
+
+echo -n '' > $OutLog
+
+alias Report 'egrep "^>" \!:1 | wc -l | awk -v P=`basename \!:1` -v H=\!:2 '"'{print H,P,"'$1}'"'"' >> $OutLog'
+
+#
+# remove entries with bad symbols
+#
+
+Notify "cleanup $ProtFile"
+
+Report $ProtFile "init_size"
+
+$AwkCmd -f $LIB_DIR/db.filter.sym.awk $ProtFile > P_$$
+
+Report $ProtFile "cleanup_size"
+
+#
+# select by name pattern
+#
+
+Notify "select by patterns"
+
+mkdir D_$$
+mkdir E_$$
+mkdir F_$$
+
+set noms = `awk '{print $1}' $PatFile`
+
+foreach nom ($noms)
+  set pat = `egrep "^$nom " $PatFile | awk '{print $2}'`
+  $AwkCmd -f $LIB_DIR/db.filter.pat.awk -v PAT="$pat" P_$$ > D_$$/$nom.fst
+  set n = `egrep '^>' D_$$/$nom.fst | wc -l`
+  Notify "  pattern : $nom : $n"
+  Report D_$$/$nom.fst "pattern_filter"
+  if ($n <= $Sizmin) \rm -f D_$$/$nom.fst
+end
+
+set ok = `ls D_$$ | wc -l`
+if ($ok == 0) goto fin
+
+#
+# select by length
+#
+
+Notify "select by length"
+
+foreach f (D_$$/*.fst) 
+  set nom = `basename $f:r`
+  $AwkCmd -f $LIB_DIR/db.getlen.awk $f > L_$$
+  $LIB_DIR/db.filter.len.r L_$$ $Delta |\
+    $AwkCmd '($NF == "TRUE") {print $2}' > M_$$
+  $AwkCmd -v FILE=M_$$ -f $LIB_DIR/db.subdb.awk $f > E_$$/$nom.fst
+  Report E_$$/$nom.fst "length_filter"
+  set n = `egrep '^>' E_$$/$nom.fst | wc -l`
+  Notify "  length filter : $nom : $n"
+  if ($n <= $Sizmin) \rm -f E_$$/$nom.fst
+end
+
+set ok = `ls E_$$ | wc -l`
+if ($ok == 0) goto fin
+
+
+#
+# select by similarity
+#
+
+Notify "select by similarity"
+
+foreach f (E_$$/*.fst) 
+  set nom = `basename $f:r`
+
+  Notify "  blasting $nom"
+  
+  makeblastdb -dbtype 'prot' -in $f >>& db.log
+  blastp -db $f -query $f -outfmt 7 > $f.blast.out
+  \rm -f $f.p??
+  
+  $AwkCmd -v COVMIN=$Covmin -v PMAX=$Pmax -v IDMIN=$Idmin \
+      -f $LIB_DIR/db.blastlink.awk $f.blast.out |\
+      $AwkCmd -f $LIB_DIR/db.todl.awk > G_$$
+      
+  ($LIB_DIR/db.cc.r G_$$ > $f.cc.txt) >>& db.log
+  
+  awk -v NAME=$nom -f $LIB_DIR/db.reportcc.awk $f.cc.txt >> $OutLog 
+  
+  $AwkCmd -f $LIB_DIR/db.selcc.awk $f.cc.txt > S_$$
+  $AwkCmd -v FILE=S_$$ -f $LIB_DIR/db.subdb.awk $f > F_$$/$nom.fst
+
+  Report F_$$/$nom.fst "similarity_filter"
+
+  set n = `egrep '^>' F_$$/$nom.fst | wc -l`
+  Notify "  blast filter : $nom : $n"
+  if ($n <= $Sizmin) \rm -f F_$$/$nom.fst
+  
+end
+
+set ok = `ls D_$$ | wc -l`
+if ($ok == 0) goto fin
+
+#
+# annotations
+#
+
+echo -n "" > J_$$
+
+foreach f (F_$$/*.fst) 
+  $AwkCmd -f $LIB_DIR/db.annot.awk $f >> J_$$
+end
+
+awk '(NF >= 3) {print $1, $NF}' $PatFile | sort > A_$$
+sort J_$$ | egrep -v '^ *$' > B_$$
+join A_$$ B_$$ > F_$$/Annot.lst
+
+#
+# copy files
+#
+
+set n = `ls F_$$/* | wc -l`
+Notify "copy $n files to $OutDir"
+
+\mv -f F_$$/* $OutDir
+
+#
+# end
+#
+
+fin:
+Notify "output directory : $OutDir"
+
+\rm -r ?_$$
+
+
+exit 0

detectors/cds/tools/chlorodb/subdb/lib/db.annot.awk

@@ -0,0 +1,39 @@
+#
+
+/^>/ {
+  N++
+  na = split($1, a, "@")
+  if (a[na-1] > NEXMAX) NEXMAX = a[na-1]
+  NEX[a[na-1]]++
+  ANNOT[$NF]++
+}
+
+END {
+  na = split(FILENAME, a, "/")
+  na = split(a[na], a, "\\.")
+  printf("%s %d ", a[1], N)
+  s = ""
+  for (i = 1 ; i <= NEXMAX ; i ++) {
+    if (NEX[i] != 0)
+      s = s "" i ":" NEX[i] "_"
+  }
+  gsub("_+$", "", s)
+  printf("%s ", s)
+  
+  s = (NEXMAX == 1) ? "MONEX" : "POLYEX"
+  printf("%s ", s)
+  
+  nmax = 0
+  amax = "none"
+  for (e in ANNOT) {
+    if (ANNOT[e] > nmax) {
+      nmax = ANNOT[e]
+      amax = e
+    }
+  }
+  print amax
+  
+}
+
+
+

detectors/cds/tools/chlorodb/subdb/lib/db.blastlink.awk

@@ -0,0 +1,48 @@
+#
+
+function min(x, y) {
+  return ((x < y) ? x : y)
+}
+
+BEGIN {
+  if (COVMIN == "") COVMIN = 50
+  if (PMAX == "")   PMAX   = 1e-6
+  if (IDMIN == "")  IDMIN  = 30
+}
+
+/^#/ {
+  hitnum = 0;
+  next;
+}
+
+{
+  if ($1 == $2) next
+  
+  hitnum++;
+  
+  na = split($1, a, "@");
+  if (na < 2) {
+    print "query file not properly formatted" > "/dev/stderr"
+    exit(1);
+  }
+  len1  = a[na];
+
+  na = split($2, a, "@");
+  if (na < 2) {
+    print "bank file not properly formatted" > "/dev/stderr"
+    exit(1);
+  }
+  len2 = a[na];
+  
+  id  = $3 + 0.0;
+  ali = $4;
+
+  covmin = ali * 100. / min(len1, len2);
+  
+  proba = $11 + 0.0;
+  
+  if ((covmin > COVMIN) && ((proba < PMAX) || (proba == 0)) && (id > IDMIN)) {
+    print $1, $2, hitnum, id, covmin, proba, ali, len1, len2;
+  }
+}
+

detectors/cds/tools/chlorodb/subdb/lib/db.cc.r

@@ -0,0 +1,18 @@
+#!/usr/bin/env Rscript
+#
+
+require(igraph, warn.conflicts=F)
+
+args <- commandArgs(T)
+path <- if(length(args) > 0) args[1] else 'graph.dl'
+
+g <- read.graph(path, format='dl')
+
+cc <- clusters(g)
+
+res <- cbind(V(g)$name, membership(cc))
+
+write.table(res, quote=FALSE, row.names=FALSE, col.names=FALSE)
+
+quit(save="no")
+

detectors/cds/tools/chlorodb/subdb/lib/db.filter.len.r

@@ -0,0 +1,19 @@
+#!/usr/bin/env Rscript
+#
+
+args <- commandArgs(T)
+path  <- if(length(args) > 0) args[1] else 'len.txt'
+delta <- if(length(args) > 1) args[2] else 0.5
+
+tab <- read.table(path, header=T)
+
+lmed <- median(tab$len)
+
+dlen <- lmed * as.numeric(delta)
+
+tab$ok <- (abs(tab$len-lmed)/lmed) <= delta
+
+write.table(tab, quote=F)
+
+quit(save='no')
+

detectors/cds/tools/chlorodb/subdb/lib/db.filter.pat.awk

@@ -0,0 +1,10 @@
+#
+
+/^>/ {
+  split($1, a, "@")
+  ok = a[3] ~ PAT 
+}
+
+ok {
+  print $0
+}

detectors/cds/tools/chlorodb/subdb/lib/db.filter.sym.awk

@@ -0,0 +1,30 @@
+#
+#
+#
+
+function Check(seq) {
+  if (seq == "") return 0
+  gsub("[ACDEFGHIKLMNPQRSTVWXY\n]+", "", seq)
+  return (length(seq) == 0)
+}
+
+/^>/ {
+  if (Check(Seq)) {
+    print Name
+    printf("%s", Seq)
+  }
+  Name = $0
+  Seq = ""
+  next
+}
+
+{
+  Seq = Seq "" $0 "\n"
+}
+
+END {
+  if (Check(Seq)) {
+    print Name
+    printf("%s", Seq)
+  }
+}

detectors/cds/tools/chlorodb/subdb/lib/db.getlen.awk

@@ -0,0 +1,10 @@
+#
+BEGIN {
+  print "id len"
+}
+
+/^>/ {
+ na = split($1, a, "@")
+ print substr($1, 2), a[na]
+}
+

detectors/cds/tools/chlorodb/subdb/lib/db.reportcc.awk

@@ -0,0 +1,15 @@
+#
+#
+
+{
+  cnt[$NF]++
+}
+
+END {
+  n = asort(cnt)
+  printf("cc_size %s", NAME)
+  for (i = n ; i >= 1 ; i--)
+    printf(" %d", cnt[i])
+  print ""
+}
+

detectors/cds/tools/chlorodb/subdb/lib/db.selcc.awk

@@ -0,0 +1,19 @@
+#
+
+{
+  N[$NF]++
+  E[$NF, N[$NF]] = $1
+}
+
+END {
+  cmax = 1
+  nmax = N[1]
+  for (i in N) {
+    if (N[i] > nmax) {
+      nmax = N[i]
+      cmax = i
+    }
+  }
+  for (i = 1 ; i <= nmax ; i++)
+    print E[cmax, i]
+}

detectors/cds/tools/chlorodb/subdb/lib/db.subdb.awk

@@ -0,0 +1,17 @@
+#
+
+BEGIN {
+  if (FILE == "") FILE = "db.sel.txt"
+  while (getline < FILE)
+    INC[$1] = $1
+  close(FILE)
+}
+
+/^>/ {
+  name = substr($1, 2)
+  ok = name in INC
+}
+
+ok {
+  print $0
+}

detectors/cds/tools/chlorodb/subdb/lib/db.todl.awk

@@ -0,0 +1,21 @@
+#
+
+{
+  node[$1]++
+  node[$2]++
+  link[++M] = $1 " " $2
+}
+
+
+END {
+ for (n in node)
+   N++
+ print "DL n=" N
+ print "format = edgelist1"
+ print "labels embedded:"
+ print "data:"
+ for (i = 1 ; i <= M ; i++)
+   print link[i]
+}
+
+