[optimize] Add parseClustersFile

assess.go

@@ -62,6 +62,7 @@ func (r *Assesser) Run() {
 	r.makeModel(r.Seqid + ".distribution.txt")
 	r.computePosteriorProb()
 	r.writePostProb(r.Seqid + ".postprob.txt")
+	log.Notice("Success")
 }
 
 // makeModel computes the norms and bins separately to derive an empirical link size

build.go

@@ -127,6 +127,7 @@ func (r *Builder) writeAGP(oo *OO) {
 func (r *Builder) Run() {
 	r.AGPfile = RemoveExt(r.Tourfile) + ".agp"
 	r.Build(r.readFiles())
+	log.Notice("Success")
 }
 
 // Build constructs molecule using component FASTA sequence

clm.go

@@ -32,7 +32,7 @@ import (
 type CLM struct {
 	REfile           string
 	Clmfile          string
-	Tigs             []TigF
+	Tigs             []*TigF
 	Tour             Tour
 	Signs            []byte
 	tigToIdx         map[string]int          // From name of the tig to the idx of the Tigs array
@@ -120,7 +120,7 @@ func (r *CLM) ParseIds() {
 		words := strings.Fields(scanner.Text())
 		tig := words[0]
 		size, _ := strconv.Atoi(words[len(words)-1])
-		r.Tigs = append(r.Tigs, TigF{idx, tig, size, true})
+		r.Tigs = append(r.Tigs, &TigF{idx, tig, size, false})
 		r.tigToIdx[tig] = idx
 		idx++
 	}

cluster.go

@@ -340,8 +340,8 @@ func (r *Partitioner) printClusters() {
 		}
 		sort.Strings(names)
 
-		fmt.Printf("g%d\t%d\t%s\n", j, len(names), strings.Join(names, ","))
-		fmt.Fprintf(w, "g%d\t%d\t%s\n", j, len(names), strings.Join(names, ","))
+		fmt.Printf("g%d\t%d\t%s\n", j, len(names), strings.Join(names, " "))
+		fmt.Fprintf(w, "g%d\t%d\t%s\n", j, len(names), strings.Join(names, " "))
 	}
 	w.Flush()
 

cmd/allhic.go

@@ -179,8 +179,8 @@ on a cluster).
 					Usage: "Skip GA step",
 				},
 				cli.BoolFlag{
-					Name:  "startOver",
-					Usage: "Do not resume from existing tour file",
+					Name:  "resume",
+					Usage: "Resume from existing tour file",
 				},
 				cli.Int64Flag{
 					Name:  "seed",
@@ -214,14 +214,14 @@ on a cluster).
 				clustersfile := c.Args().Get(2)
 				group, _ := strconv.Atoi(c.Args().Get(3))
 				runGA := !c.Bool("skipGA")
-				startOver := c.Bool("startOver")
+				resume := c.Bool("resume")
 				seed := c.Int64("seed")
 				npop := c.Int("npop")
 				ngen := c.Int("ngen")
 				mutpb := c.Float64("mutpb")
 				p := allhic.Optimizer{REfile: refile, Clmfile: clmfile,
 					Clustersfile: clustersfile, Group: group,
-					RunGA: runGA, StartOver: startOver,
+					RunGA: runGA, Resume: resume,
 					Seed: seed, NPop: npop, NGen: ngen, MutProb: mutpb}
 				p.Run()
 				return nil

extract.go

@@ -133,6 +133,7 @@ func (r *Extracter) Run() {
 	r.makeModel(RemoveExt(r.Bamfile) + ".distribution.txt")
 	r.calcIntraContigs()
 	r.calcInterContigs()
+	log.Notice("Success")
 }
 
 // makeModel computes the norms and bins separately to derive an empirical link size

optimize.go

@@ -24,7 +24,7 @@ type Optimizer struct {
 	Clustersfile string
 	Group        int
 	RunGA        bool
-	StartOver    bool
+	Resume       bool
 	Seed         int64
 	NPop         int
 	NGen         int
@@ -35,25 +35,25 @@ type Optimizer struct {
 // Run kicks off the Optimizer
 func (r *Optimizer) Run() {
 	clm := NewCLM(r.Clmfile, r.REfile)
-	tourfile := RemoveExt(r.Clmfile) + ".tour"
-	shuffle := false
+	tourfile := fmt.Sprintf("%s.g%d.tour", RemoveExt(r.Clmfile), r.Group)
 
 	// Load tourfile if it exists
-	if _, err := os.Stat(tourfile); !r.StartOver && err == nil {
-		log.Noticef("Found existing tour file")
+	if _, err := os.Stat(tourfile); r.Resume && err == nil {
+		log.Noticef("Found existing tour file `%s`", tourfile)
 		clm.parseTourFile(tourfile)
-		clm.printTour(os.Stdout, clm.Tour, "INIT")
 		// Rename the tour file
 		backupTourFile := tourfile + ".sav"
 		os.Rename(tourfile, backupTourFile)
 		log.Noticef("Backup `%s` to `%s`", tourfile, backupTourFile)
 	} else {
-		shuffle = true
+		clm.parseClustersFile(r.Clustersfile, r.Group)
 	}
 
+	shuffle := false // If one wants randomized initialization, set this to true
 	clm.Activate(shuffle)
 
 	// tourfile logs the intermediate configurations
+	log.Noticef("Optimization history logged to `%s`", tourfile)
 	fwtour, _ := os.Create(tourfile)
 	defer fwtour.Close()
 	clm.printTour(fwtour, clm.Tour, "INIT")
@@ -112,16 +112,21 @@ func parseTourFile(filename string) []string {
 	return words
 }
 
-// parseTourFile parses tour file
-// Only the last line is retained anc onverted into a Tour
-func (r *CLM) parseTourFile(filename string) {
-	words := parseTourFile(filename)
-	tigs := []Tig{}
+// prepareTour prepares a boilerplate for an empty tour
+func (r *CLM) prepareTour() {
 	r.Signs = make([]byte, len(r.Tigs))
 	for _, tig := range r.Tigs {
 		tig.IsActive = false
 	}
+}
 
+// parseTourFile parses tour file
+// Only the last line is retained anc onverted into a Tour
+func (r *CLM) parseTourFile(filename string) {
+	words := parseTourFile(filename)
+	r.prepareTour()
+
+	tigs := []Tig{}
 	for _, word := range words {
 		tigName, tigOrientation := word[:len(word)-1], word[len(word)-1]
 		idx, ok := r.tigToIdx[tigName]
@@ -133,9 +138,34 @@ func (r *CLM) parseTourFile(filename string) {
 			Idx: idx,
 		})
 		r.Signs[idx] = tigOrientation
-		r.Tour.Tigs = tigs
 		r.Tigs[idx].IsActive = true
 	}
+	r.Tour.Tigs = tigs
+	r.printTour(os.Stdout, r.Tour, "INIT")
+}
+
+// parseClustersFile parses clusters file
+func (r *CLM) parseClustersFile(clustersfile string, group int) {
+	recs := ReadCSVLines(clustersfile)
+	r.prepareTour()
+
+	rec := recs[group]
+	names := strings.Split(rec[2], " ")
+	tigs := []Tig{}
+	for _, tigName := range names {
+		idx, ok := r.tigToIdx[tigName]
+		if !ok {
+			log.Errorf("Contig %s not found!", tigName)
+			continue
+		}
+		tigs = append(tigs, Tig{
+			Idx: idx,
+		})
+		r.Signs[idx] = '+'
+		r.Tigs[idx].IsActive = true
+	}
+	r.Tour.Tigs = tigs
+	r.printTour(os.Stdout, r.Tour, "INIT")
 }
 
 // printTour logs the current tour to file

partition.go

@@ -175,7 +175,7 @@ func (r *Partitioner) readRE() {
 		}
 		r.contigs = append(r.contigs, ci)
 	}
-	r.contigToIdx = make(map[string]int)
+	r.contigToIdx = map[string]int{}
 	for i, contig := range r.contigs {
 		r.contigToIdx[contig.name] = i
 	}