static int ReadMotifPass2(FILE *f, MotifData *Motifs)
{
rewind(f);
GFFLineNr = 0;
int MotifCount = 0;
GFFRecord Rec;
while (GetNextGFFRecord(f, Rec))
{
if (0 != strcmp(Rec.Feature, "tandemmotif"))
continue;
if (Rec.Start <= 0 || Rec.End <= 0 || Rec.Start > Rec.End)
Warning("GFF line %d: invalid start %d / end %d",
GFFLineNr, Rec.Start, Rec.End);
int FamIndex = GetFam(Rec);
MotifData &Motif = Motifs[MotifCount];
const int Length = Rec.End - Rec.Start + 1;
Motif.ContigLabel = strsave(Rec.SeqName);
Motif.ContigFrom = Rec.Start - 1;
Motif.ContigTo = Motif.ContigFrom + Length - 1;
Motif.FamIndex = FamIndex;
++MotifCount;
}
return MotifCount;
}
void GFFSet::FromFile(FILE *f)
{
Free();
GFFRecord Rec;
while (GetNextGFFRecord(f, Rec))
{
SaveGFFStrings(Rec);
Add(Rec);
}
}
static int ReadTRSPass1(FILE *f)
{
GFFLineNr = 0;
int TRSCount = 0;
GFFRecord Rec;
while (GetNextGFFRecord(f, Rec))
{
if (0 != strcmp(Rec.Feature, "trs"))
continue;
if (Rec.Start <= 0 || Rec.End <= 0 || Rec.Start > Rec.End)
Warning("GFF line %d: invalid start %d / end %d",
GFFLineNr, Rec.Start, Rec.End);
++TRSCount;
}
return TRSCount;
}
static int ReadTRSPass2(FILE *f, TRSData *TRSs)
{
rewind(f);
GFFLineNr = 0;
int TRSCount = 0;
GFFRecord Rec;
while (GetNextGFFRecord(f, Rec))
{
if (0 != strcmp(Rec.Feature, "trs"))
continue;
if (Rec.Start <= 0 || Rec.End <= 0 || Rec.Start > Rec.End)
Warning("GFF line %d: invalid start %d / end %d",
GFFLineNr, Rec.Start, Rec.End);
static char *Fam = GetFam(Rec);
int FamIndex = 0;
int SuperFamIndex = 0;
int n = sscanf(Fam, "%d.%d", &FamIndex, &SuperFamIndex);
if (n == 1)
SuperFamIndex = -1;
else if (n != 2)
Quit("Invalid Family %s", Fam);
TRSData &TRS = TRSs[TRSCount];
const int Length = Rec.End - Rec.Start + 1;
TRS.ContigLabel = strsave(Rec.SeqName);
TRS.ContigFrom = Rec.Start - 1;
TRS.ContigTo = TRS.ContigFrom + Length - 1;
TRS.FamIndex = FamIndex;
TRS.SuperFamIndex = SuperFamIndex;
if (Rec.Strand == '+')
TRS.Rev = false;
else if (Rec.Strand == '-')
TRS.Rev = true;
else
Quit("GFF line %d, Invalid strand %c", GFFLineNr, Rec.Strand);
++TRSCount;
}
return TRSCount;
}
void AnnotEdge()
{
const char *InputFileName = RequiredValueOpt("annotedge");
const char *RepeatFileName = RequiredValueOpt("rep");
const char *OutputFileName = RequiredValueOpt("out");
ProgressStart("Reading repeat file");
int RepCount;
RepData *Reps = ReadReps(RepeatFileName, &RepCount);
ProgressDone();
Progress("%d records", RepCount);
FILE *fInput = OpenStdioFile(InputFileName);
FILE *fOutput = OpenStdioFile(OutputFileName, FILEIO_MODE_WriteOnly);
ProgressStart("Transferring annotation");
GFFRecord Rec;
while (GetNextGFFRecord(fInput, Rec))
{
const bool Rev = (Rec.Strand == '-');
const char *Annot = MakeAnnotEdge(Rec.SeqName, Rec.Start-1, Rec.End-1, Rev,
Reps, RepCount);
fprintf(fOutput, "%s\t%s\t%s\t%d\t%d\t%.3g\t%c",
// 0 1 2 3 4 5 6
Rec.SeqName, // 0
Rec.Source, // 1
Rec.Feature, // 2
Rec.Start, // 3
Rec.End, // 4
Rec.Score, // 5
Rec.Strand); // 6
if (-1 == Rec.Frame)
fprintf(fOutput, "\t.");
else
fprintf(fOutput, "\t%d", Rec.Frame);
fprintf(fOutput, "\t%s ; Annot \"%s\"\n", Rec.Attrs, Annot);
}
fclose(fInput);
fclose(fOutput);
ProgressDone();
}
int GLIX::FromGFFFile(FILE *f)
{
int RecordCount = 0;
GFFRecord Rec;
while (GetNextGFFRecord(f, Rec))
{
++RecordCount;
Add(Rec.SeqName, Rec.End - 1);
if (HasTargetAttrs(Rec.Attrs))
{
char TargetName[MAX_GFF_FEATURE_LENGTH+1];
int Start;
int End;
ParseTargetAttrs(Rec.Attrs, TargetName, sizeof(TargetName), &Start, &End);
Add(TargetName, End - 1);
}
}
AssignOffsets();
return RecordCount;
}
static int ReadRepsPass2(FILE *f, RepData *Reps)
{
rewind(f);
GFFLineNr = 0;
int RepCount = 0;
GFFRecord Rec;
while (GetNextGFFRecord(f, Rec))
{
if (0 != strcmp(Rec.Feature, "repeat"))
continue;
static char *Repeat = GetRepeat(Rec);
RepData &Rep = Reps[RepCount];
ParseRepeat(Repeat, Rep);
if (Rec.Start <= 0 || Rec.End <= 0 || Rec.Start > Rec.End)
Warning("GFF line %d: invalid start %d / end %d",
GFFLineNr, Rec.Start, Rec.End);
const int Length = Rec.End - Rec.Start + 1;
Rep.ContigLabel = strsave(Rec.SeqName);
Rep.ContigFrom = Rec.Start - 1;
Rep.ContigTo = Rep.ContigFrom + Length - 1;
if (Rec.Strand == '+')
Rep.Rev = false;
else if (Rec.Strand == '-')
Rep.Rev = true;
else
Quit("GFF line %d, Invalid strand %c", GFFLineNr, Rec.Strand);
++RepCount;
}
return RepCount;
}
static int ReadRepsPass1(FILE *f)
{
GFFLineNr = 0;
int RepCount = 0;
GFFRecord Rec;
while (GetNextGFFRecord(f, Rec))
{
if (0 != strcmp(Rec.Feature, "repeat"))
{
static bool WarningIssued = false;
if (!WarningIssued)
{
Warning("GFF record feature '%s' is not a repeat", Rec.Feature);
WarningIssued = true;
}
continue;
}
if (Rec.Start <= 0 || Rec.End <= 0 || Rec.Start > Rec.End)
Warning("GFF line %d: invalid start %d / end %d",
GFFLineNr, Rec.Start, Rec.End);
++RepCount;
}
return RepCount;
}
void Tan()
{
// Image file annotated with from-to pile indexes
// Produced by:
// piler2 -trs banded_hits.gff -images mainband_images.gff
const char *HitFileName = RequiredValueOpt("tan");
const char *OutFileName = RequiredValueOpt("out");
const char *PyramidFileName = ValueOpt("pyramid");
const char *MotifFileName = ValueOpt("motif");
const char *strMinHits = ValueOpt("minhits");
const char *strMaxMargin = ValueOpt("maxmargin");
const char *strMinRatio = ValueOpt("minratio");
if (0 != strMinHits)
MIN_HIT_COUNT = atoi(strMinHits);
if (0 != strMaxMargin)
MAX_FRACT_MARGIN = atof(strMaxMargin);
if (0 != strMinRatio)
MIN_RATIO = atof(strMinRatio);
FILE *fInput = OpenStdioFile(HitFileName);
ProgressStart("Initialize piles");
GFFRecord Rec;
int HitCount = 0;
while (GetNextGFFRecord(fInput, Rec))
{
if (0 != strcmp(Rec.Feature, "hit"))
continue;
int QueryPileIndex = -1;
int TargetPileIndex = -1;
ParsePilesAttrs(Rec.Attrs, &QueryPileIndex, &TargetPileIndex);
if (QueryPileIndex != TargetPileIndex)
continue;
char TargetLabel[128];
int TargetStart;
int TargetEnd;
ParseTargetAttrs(Rec.Attrs, TargetLabel, sizeof(TargetLabel), &TargetStart, &TargetEnd);
if (0 != strcmp(Rec.SeqName, TargetLabel))
Quit("Labels don't match");
const int QueryFrom = Rec.Start - 1;
const int QueryTo = Rec.End - 1;
const int TargetFrom = TargetStart - 1;
const int TargetTo = TargetEnd - 1;
const bool Rev = (Rec.Strand == '-');
AddHit(QueryPileIndex, Rec.SeqName, QueryFrom, QueryTo, TargetFrom, TargetTo, Rev);
++HitCount;
}
ProgressDone();
Progress("%d hits, %d piles", HitCount, PileCount);
ProgressStart("Allocate piles");
for (int PileIndex = 0; PileIndex < PileCount; ++PileIndex)
{
TanPile &Pile = Piles[PileIndex];
Pile.Hits = all(HitData, Pile.HitCount);
Pile.HitCount = 0;
}
ProgressDone();
ProgressStart("Assign hits to piles");
Rewind(fInput);
while (GetNextGFFRecord(fInput, Rec))
{
if (0 != strcmp(Rec.Feature, "hit"))
continue;
int QueryPileIndex = -1;
int TargetPileIndex = -1;
ParsePilesAttrs(Rec.Attrs, &QueryPileIndex, &TargetPileIndex);
if (QueryPileIndex != TargetPileIndex)
continue;
char TargetLabel[128];
int TargetStart;
int TargetEnd;
ParseTargetAttrs(Rec.Attrs, TargetLabel, sizeof(TargetLabel), &TargetStart, &TargetEnd);
if (0 != strcmp(Rec.SeqName, TargetLabel))
Quit("Labels don't match");
const int QueryFrom = Rec.Start - 1;
const int QueryTo = Rec.End - 1;
const int TargetFrom = TargetStart - 1;
const int TargetTo = TargetEnd - 1;
const bool Rev = (Rec.Strand == '-');
AssignHit(QueryPileIndex, Rec.SeqName, QueryFrom, QueryTo, TargetFrom, TargetTo, Rev);
}
ProgressDone();
fOut = OpenStdioFile(OutFileName, FILEIO_MODE_WriteOnly);
fPyramid = (0 == PyramidFileName ? 0 : OpenStdioFile(PyramidFileName, FILEIO_MODE_WriteOnly));
fMotif = (0 == PyramidFileName ? 0 : OpenStdioFile(MotifFileName, FILEIO_MODE_WriteOnly));
ProgressStart("Find pyramids");
for (int PileIndex = 0; PileIndex < PileCount; ++PileIndex)
FindPyramids(PileIndex);
int PyramidCount = PyramidIndex;
ProgressDone();
Progress("%d pyramids", PyramidCount);
}
void TR()
{
#if defined(DEBUG) && defined(_MSC_VER)
_CrtSetDbgFlag(0); // too expensive
#endif
const char *HitFileName = RequiredValueOpt("tr");
const char *OutFileName = RequiredValueOpt("out");
const char *CandFileName = ValueOpt("cand");
const char *strMinTrSpacing = ValueOpt("mintrspacing");
const char *strMaxTrSpacing = ValueOpt("maxtrspacing");
const char *strMinTrLength = ValueOpt("mintrlength");
const char *strMaxTrLength = ValueOpt("minspacingratio");
const char *strMinFam = ValueOpt("minfam");
const char *strMinHitRatio = ValueOpt("minhitratio");
const char *strMinDistPairs = ValueOpt("mindistpairs");
if (0 != strMinTrSpacing)
MIN_LENGTH_LINE = atoi(strMinTrSpacing);
if (0 != strMaxTrSpacing)
MAX_LENGTH_LINE = atoi(strMaxTrSpacing);
if (0 != strMinTrLength)
MIN_LENGTH_LTR = atoi(strMinTrLength);
if (0 != strMaxTrLength)
MAX_LENGTH_LTR = atoi(strMaxTrLength);
if (0 != strMinFam)
MIN_FAM_SIZE = atoi(strMinFam);
if (0 != strMinHitRatio)
MIN_HIT_LENGTH_RATIO = atoi(strMinHitRatio);
if (0 != strMinDistPairs)
MIN_DIST_EDGE = atoi(strMinDistPairs);
FILE *fHit = OpenStdioFile(HitFileName, FILEIO_MODE_ReadOnly);
ProgressStart("Index hits");
GLIX HitGlix;
HitGlix.Init();
HitGlix.FromGFFFile(fHit);
HitGlix.MakeGlobalToLocalIndex();
ProgressDone();
const int GlobalLength = HitGlix.GetGlobalLength();
IIX IntervalIndex;
IntervalIndex.Init(GlobalLength);
ProgressStart("Find candidate TRs");
Rewind(fHit);
GFFRecord Rec;
while (GetNextGFFRecord(fHit, Rec))
{
HitData Hit;
GFFRecordToHit(HitGlix, Rec, Hit);
if (IsCandLTR(Hit))
AddCand(Hit, IntervalIndex);
}
ProgressDone();
Progress("%d candidates", CandCount);
if (0 != CandFileName)
{
ProgressStart("Write candidates");
FILE *fCand = OpenStdioFile(CandFileName, FILEIO_MODE_WriteOnly);
WriteCands(fCand, HitGlix);
ProgressDone();
}
ProgressStart("Make graph");
Rewind(fHit);
while (GetNextGFFRecord(fHit, Rec))
{
HitData Hit;
GFFRecordToHit(HitGlix, Rec, Hit);
FindEdges(Hit, HitGlix, IntervalIndex);
}
fclose(fHit);
fHit = 0;
ProgressDone();
Progress("%d edges", (int) Edges.size());
ProgressStart("Find families");
FamList Fams;
FindConnectedComponents(Edges, Fams, MIN_FAM_SIZE);
ProgressDone();
Progress("%d families", (int) Fams.size());
FILE *fOut = OpenStdioFile(OutFileName, FILEIO_MODE_WriteOnly);
WriteOutputFile(fOut, HitGlix, Fams);
}
