Added code for building sets of primers and -p command line option

git-svn-id: https://www.grenoble.prabi.fr/svn/LECASofts/ecoPrimers/trunk@279 60f365c0-8329-0410-b2a4-ec073aeeaa1d

src/ecoprimer.c

@@ -6,6 +6,7 @@
  */
 
 #include "libecoprimer/ecoprimer.h"
+#include "libecoprimer/PrimerSets.h"
 #include <stdio.h>
 #include <string.h>
 #include <ctype.h>
@@ -74,6 +75,7 @@ static void PrintHelp()
               PP      "-A    : Print the list of all identifier of sequences present in the database\n\n");
               PP      "-f    : Remove data mining step during  strict primer identification\n\n");
               PP      "-v    : Store statistic file about memory usage during strict primer identification\n\n");
+              PP      "-p    : Print sets of primers\n\n");
               PP      "\n");
               PP      "------------------------------------------\n");
               PP      "Table result description : \n");
@@ -132,9 +134,9 @@ void initoptions(poptions_t options)
 	options->refseq=NULL;
 	options->circular=0;
 	options->doublestrand=1;
-	options->strict_quorum=0.7;
+	options->strict_quorum=0.3;
 	options->strict_exclude_quorum=0.1;
-	options->sensitivity_quorum=0.9;
+	options->sensitivity_quorum=0.3;
 	options->false_positive_quorum=0.1;
 	options->strict_three_prime=0;
 	options->r=0;
@@ -146,6 +148,7 @@ void initoptions(poptions_t options)
 	options->saltmethod = SALT_METHOD_SANTALUCIA;
 	options->salt = DEF_SALT;
 	options->printAC=FALSE;
+	options->print_sets_of_primers = FALSE;
 }
 
 void printapair(int32_t index,ppair_t pair, poptions_t options)
@@ -314,7 +317,7 @@ static int cmpprintedpairs(const void* p1,const void* p2)
 	return 0;
 }
 
-uint32_t filterandsortpairs(ppair_t* sortedpairs,uint32_t count, poptions_t options)
+uint32_t filterandsortpairs(ppair_t* sortedpairs,uint32_t count, poptions_t options, pecodnadb_t seqdb)
 {
 	uint32_t i,j;
 	float q,qfp;
@@ -329,6 +332,7 @@ uint32_t filterandsortpairs(ppair_t* sortedpairs,uint32_t count, poptions_t opti
 			qfp = (float)sortedpairs[i]->outexample/options->outsamples;
 		else qfp=0.0;
 
+		sortedpairs[i]->wellIdentifiedSeqs = NULL; //TR 05/09/10 - wellIdentified needed for primer sets
 		sortedpairs[i]->quorumin = q;
 		sortedpairs[i]->quorumout = qfp;
 		sortedpairs[i]->yule = q - qfp;
@@ -337,8 +341,10 @@ uint32_t filterandsortpairs(ppair_t* sortedpairs,uint32_t count, poptions_t opti
 		if (q > options->sensitivity_quorum &&
 			qfp < options->false_positive_quorum)
 		{
+			//TR 05/09/10 - wellIdentified needed for primer sets
+			sortedpairs[j]->wellIdentifiedSeqs = ECOMALLOC(options->dbsize * sizeof(int),"Cannot allocate well_identified_array");
 			(void)taxonomycoverage(sortedpairs[j],options);
-			taxonomyspecificity(sortedpairs[j]);
+			taxonomyspecificity(sortedpairs[j], seqdb, options->dbsize);
 			j++;
 		}
 
@@ -348,7 +354,7 @@ uint32_t filterandsortpairs(ppair_t* sortedpairs,uint32_t count, poptions_t opti
 }
 
 
-void printpairs (ppairtree_t pairs, poptions_t options,ecotaxonomy_t *taxonomy)
+void printpairs (ppairtree_t pairs, poptions_t options,ecotaxonomy_t *taxonomy, pecodnadb_t seqdb)
 {
    ppair_t* sortedpairs;
    ppair_t* index;
@@ -357,7 +363,7 @@ void printpairs (ppairtree_t pairs, poptions_t options,ecotaxonomy_t *taxonomy)
    size_t count;
    char *taxon[]={"taxon","taxa"};
    ecotx_t    *current_taxon;
-
+   pairset pair_sets;
 
    //printf("Index\tPrimer1\tPrimer2\tGB\tInexampleCount\tOutexampleCount\tYule\tIntaxaCount\tOuttaxaCount\tCoverage\tSpecificity\tMinAmplifiedLength\tMaxAmplifiedLength\tAvgAmplifiedLength\n");
 
@@ -377,7 +383,7 @@ void printpairs (ppairtree_t pairs, poptions_t options,ecotaxonomy_t *taxonomy)
    for (i=0;i<pl->paircount;i++,j++)
 	   sortedpairs[j]=pl->pairs+i;
 
-   count=filterandsortpairs(sortedpairs,pairs->count,options);
+   count=filterandsortpairs(sortedpairs,pairs->count,options, seqdb);
    getThermoProperties(sortedpairs, count, options);
 
    fprintf(stderr,"Total good pair count : %u\n",(uint32_t)count);
@@ -441,8 +447,12 @@ void printpairs (ppairtree_t pairs, poptions_t options,ecotaxonomy_t *taxonomy)
 
    for (i=0;i < count;i++)
 	   printapair(i,sortedpairs[i],options);
-
-
+   
+   if (options->print_sets_of_primers == TRUE)
+   {
+	   pair_sets = build_primers_set (sortedpairs, count, seqdb, options);
+	   some_other_set_possibilities (&pair_sets, sortedpairs, count, seqdb, options);
+   }
 }
 
 
@@ -528,7 +538,7 @@ int main(int argc, char **argv)
 
 	initoptions(&options);
 
-    while ((carg = getopt(argc, argv, "hAfvcUDSE:d:l:L:e:i:r:R:q:3:s:x:t:O:m:a:")) != -1) {
+    while ((carg = getopt(argc, argv, "hAfvcUDSpE:d:l:L:e:i:r:R:q:3:s:x:t:O:m:a:")) != -1) {
 
      switch (carg) {
 								 /* ---------------------------- */
@@ -691,6 +701,12 @@ int main(int argc, char **argv)
 		  options.circular = 1;
           break;
 
+								/* -------------------- */
+        case 'p':               /* print sets of primers */
+								/* --------------------------------- */
+        	options.print_sets_of_primers = TRUE;
+        break;
+
         case '?':               /* bad option           */
                                 /* -------------------- */
             errflag++;
@@ -799,9 +815,9 @@ int main(int argc, char **argv)
 
     fprintf(stderr,"\n");
 
-    pairs = buildPrimerPairs(seqdb, seqdbsize, primers, &options);
-
-     printpairs (pairs, &options,taxonomy);
-
+   
+    	pairs = buildPrimerPairs(seqdb, seqdbsize, primers, &options);
+    	printpairs (pairs, &options,taxonomy, seqdb);
+    
     return 0;
 }

src/libecoprimer/Makefile

@@ -14,7 +14,8 @@ SOURCES = goodtaxon.c \
           pairs.c \
           taxstats.c \
           apat_search.c \
-		  filtering.c
+		  filtering.c \
+		  PrimerSets.c
 
 SRCS=$(SOURCES)
          

src/libecoprimer/PrimerSets.h

+#ifndef PRIMERSETS_H_
+#define PRIMERSETS_H_
+
+#include "ecoprimer.h"
+
+#define PRIMERS_IN_SET_COUNT 10
+
+typedef struct {
+	int		*set_wellIdentifiedTaxa;
+	int32_t set_pairs[PRIMERS_IN_SET_COUNT];
+	float set_specificity;
+	float set_coverage;
+	float set_lmean;
+	float set_lcov;
+	float set_score;
+}pairset;
+
+typedef struct{
+	ppair_t* sortedpairs;
+	int32_t sorted_count;
+	pecodnadb_t seqdb;
+	poptions_t options;
+}SetParams;
+
+typedef struct{
+	float t_spc; //specificity contribution
+	float t_cov; //coverage contribution
+	float t_lmd; //link spread difference
+	float len; //length
+	float score; //score
+}primerscore;
+
+void add_pair_in_set (pairset *pair_set, int32_t pset_idx, int32_t prb_idx, SetParams *pparams);
+void get_next_pair_options (int *pair_wi_count_sorted_ids, pairset *pair_set, SetParams *pparams);
+float get_links_distribution (int prb_idx, pairset *prob_set, SetParams *pparams);
+pairset build_primers_set (ppair_t* sortedpairs, int32_t sorted_count, pecodnadb_t seqdb,
+		poptions_t options);
+void get_set_mean_cov_stats (pairset *prob_set, SetParams *pparams);
+void some_other_set_possibilities (pairset *pair_set,
+		ppair_t * sortedpairs, int32_t sorted_count, pecodnadb_t seqdb, poptions_t options);
+#endif 

src/libecoprimer/ecoprimer.h

@@ -173,6 +173,9 @@ typedef struct {
 	uint32_t   		outtaxa;            //< counterexample taxa count
 	uint32_t		notwellidentifiedtaxa;
 
+	int				*wellIdentifiedSeqs; //< an array having elements equla to total seqs
+									 // values are either 0 or 1, if seq is well identified
+									 // its 1 else 0
 
 					// these statistics are relative to inexample sequences
 
@@ -286,6 +289,7 @@ typedef struct {
 	int			  saltmethod;
 	float		  salt;
 	PNNParams	  pnparm;
+	bool_t		  print_sets_of_primers;
 } options_t, *poptions_t;
 
 typedef ecoseq_t  **pecodnadb_t;
@@ -347,7 +351,7 @@ int32_t getrankdbstats(pecodnadb_t seqdb,
 					   poptions_t options);
 float taxonomycoverage(ppair_t pair, poptions_t options);
 char ecoComplementChar(char base);
-void taxonomyspecificity (ppair_t pair);
+void taxonomyspecificity (ppair_t pair, pecodnadb_t seqdb,uint32_t seqdbsize);
 
 int32_t *filteringSeq(pecodnadb_t database, uint32_t seqdbsize,
 					 uint32_t exampleCount,poptions_t options,uint32_t *size,int32_t  sequenceQuorum);

src/libecoprimer/taxstats.c

@@ -179,9 +179,16 @@ void twalkaction (const void *node, VISIT order, int level)
 	counttaxon(taxid);
 }
 
-void taxonomyspecificity (ppair_t pair)
+int32_t gtxid;
+void twalkaction2 (const void *node, VISIT order, int level)
 {
-	uint32_t i;
+	int32_t *pt = (int32_t *) node;
+	gtxid = *pt;
+}
+
+void taxonomyspecificity (ppair_t pair, pecodnadb_t seqdb,uint32_t seqdbsize)
+{
+	uint32_t i, j;
 	uint32_t ampfindex = 0;
 	int32_t taxid;
 	void *ampftree = NULL;
@@ -219,11 +226,31 @@ void taxonomyspecificity (ppair_t pair)
 		}
 	}
 
+	memset (pair->wellIdentifiedSeqs, 0, seqdbsize*sizeof (int));
 	counttaxon(-1);
 	for (i = 0; i < ampfindex; i++)
 	{
 		if (ampfwithtaxtree[i].taxoncount > 1)
 			twalk(ampfwithtaxtree[i].taxontree, twalkaction);
+		//TR 5/9/10 - added code for well identified seqs
+		else if(ampfwithtaxtree[i].taxoncount == 1) /*well identified*/
+		{
+			gtxid = -1;
+			twalk(ampfwithtaxtree[i].taxontree, twalkaction2);
+			
+			if (gtxid != -1)
+			{
+				for (j = 0; j < seqdbsize; j++)
+					if (seqdb[j]->ranktaxonid == gtxid 
+							&&(pair->p1->directCount[j] > 0 
+							|| pair->p1->reverseCount[j] > 0)
+							&& (pair->p2->directCount[j] > 0 
+							|| pair->p2->reverseCount[j] > 0))
+					{
+						pair->wellIdentifiedSeqs[j] = 1;
+					}
+			}
+		}
 	}
 
 	pair->notwellidentifiedtaxa = counttaxon(-2);