void TRS2Fasta()
{
const char *TRSFileName = RequiredValueOpt("trs2fasta");
const char *SeqFileName = RequiredValueOpt("seq");
const char *Path = ValueOpt("path");
const char *strMaxFam = ValueOpt("maxfam");
const char *Prefix = ValueOpt("prefix");
int MaxFam = DEFAULT_MAX_FAM;
if (strMaxFam != 0)
MaxFam = atoi(strMaxFam);
if (0 == Path)
Path = ".";
ProgressStart("Reading seq file");
int SeqLength;
const char *Seq = ReadMFA(SeqFileName, &SeqLength);
ProgressDone();
Progress("Seq length %d bases, %.3g Mb", SeqLength, SeqLength/1e6);
ProgressStart("Read TRS file");
int TRSCount;
TRSData *TRSs = ReadTRS(TRSFileName, &TRSCount);
ProgressDone();
Progress("%d records", TRSCount);
ProgressStart("Sorting by family");
qsort((void *) TRSs, TRSCount, sizeof(TRSData), CmpTRS);
ProgressDone();
FILE *f = 0;
int CurrentFamily = -1;
int MemberCount = 0;
for (int TRSIndex = 0; TRSIndex < TRSCount; ++TRSIndex)
{
const TRSData &TRS = TRSs[TRSIndex];
if (TRS.FamIndex != CurrentFamily)
{
if (f != 0)
fclose(f);
char *FastaFileName = FamFileName(Path, TRS.FamIndex, TRS.SuperFamIndex);
f = OpenStdioFile(FastaFileName, FILEIO_MODE_WriteOnly);
CurrentFamily = TRS.FamIndex;
MemberCount = 0;
}
++MemberCount;
if (MemberCount > MaxFam)
continue;
const int From = ContigToGlobal(TRS.ContigFrom, TRS.ContigLabel);
const int Length = TRS.ContigTo - TRS.ContigFrom + 1;
char *Label = TRSLabel(Prefix, TRS);
WriteFasta(f, Seq + From, Length, Label, TRS.Rev);
freemem(Label);
}
}
void Tanmotif2Fasta()
{
const char *MotifFileName = RequiredValueOpt("tanmotif2fasta");
const char *SeqFileName = RequiredValueOpt("seq");
const char *Path = ValueOpt("path");
const char *strMaxFam = ValueOpt("maxfam");
const char *Prefix = ValueOpt("prefix");
int MaxFam = DEFAULT_MAX_FAM;
if (strMaxFam != 0)
MaxFam = atoi(strMaxFam);
if (0 == Path)
Path = ".";
ProgressStart("Reading seq file");
int SeqLength;
const char *Seq = ReadMFA(SeqFileName, &SeqLength);
ProgressDone();
Progress("Seq length %d bases, %.3g Mb", SeqLength, SeqLength/1e6);
ProgressStart("Read Motif file");
int MotifCount;
MotifData *Motifs = ReadMotif(MotifFileName, &MotifCount);
ProgressDone();
Progress("%d records", MotifCount);
ProgressStart("Sorting by family");
qsort((void *) Motifs, MotifCount, sizeof(MotifData), CmpMotif);
ProgressDone();
FILE *f = 0;
int CurrentFamily = -1;
int MemberCount = 0;
for (int MotifIndex = 0; MotifIndex < MotifCount; ++MotifIndex)
{
const MotifData &Motif = Motifs[MotifIndex];
if (Motif.FamIndex != CurrentFamily)
{
if (f != 0)
fclose(f);
char *FastaFileName = FamFileName(Path, Motif.FamIndex);
f = OpenStdioFile(FastaFileName, FILEIO_MODE_WriteOnly);
CurrentFamily = Motif.FamIndex;
MemberCount = 0;
}
++MemberCount;
if (MemberCount > MaxFam)
continue;
const int From = ContigToGlobal(Motif.ContigFrom, Motif.ContigLabel);
const int Length = Motif.ContigTo - Motif.ContigFrom + 1;
char *Label = MotifLabel(Prefix, Motif);
WriteFasta(f, Seq + From, Length, Label, false);
freemem(Label);
}
}
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();
}
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);
}
void TRS()
{
const char *InputFileName = RequiredValueOpt("trs");
const char *OutputFileName = ValueOpt("out");
const char *PilesFileName = ValueOpt("piles");
const char *ImagesFileName = ValueOpt("images");
const char *strMinFamSize = ValueOpt("famsize");
const char *strMaxLengthDiffPct = ValueOpt("maxlengthdiffpct");
g_paramSingleHitCoverage = !FlagOpt("multihit");
if (0 == OutputFileName && 0 == PilesFileName && 0 == ImagesFileName)
Quit("No output file specified, must be at least one of -out, -piles, -images");
if (0 != strMinFamSize)
g_paramMinFamSize = atoi(strMinFamSize);
if (0 != strMaxLengthDiffPct)
g_paramMaxLengthDiffPct = atoi(strMaxLengthDiffPct);
Log("singlehit=%s famsize=%d maxlengthdiffpct=%d\n",
g_paramSingleHitCoverage ? "True" : "False",
g_paramMinFamSize,
g_paramMaxLengthDiffPct);
ProgressStart("Read hit file");
int HitCount;
int SeqLength;
HitData *Hits = ReadHits(InputFileName, &HitCount, &SeqLength);
ProgressDone();
Progress("%d hits", HitCount);
SeqLengthChunks = (SeqLength + CHUNK_LENGTH - 1)/CHUNK_LENGTH;
const int BitVectorLength = (SeqLengthChunks + BITS_PER_INT - 1)/BITS_PER_INT;
int *CopyCount = all(int, BitVectorLength);
zero(CopyCount, int, BitVectorLength);
ProgressStart("Compute copy counts");
for (int i = 0; i < HitCount; ++i)
IncCopyCount(CopyCount, Hits[i]);
ProgressDone();
ProgressStart("Identify piles");
PILE_INDEX_TYPE *PileIndexes = IdentifyPiles(CopyCount);
ProgressDone();
Progress("%d stacks", PileCount);
freemem(CopyCount);
CopyCount = 0;
CreatePiles(Hits, HitCount, PileIndexes);
if (0 != ImagesFileName)
{
ProgressStart("Writing images file");
WriteImages(ImagesFileName, Hits, HitCount, PileIndexes);
ProgressDone();
}
freemem(Hits);
Hits = 0;
if (0 != PilesFileName)
{
ProgressStart("Writing piles file");
WritePiles(PilesFileName, Piles, PileCount);
ProgressDone();
}
freemem(PileIndexes);
PileIndexes = 0;
if (0 == OutputFileName)
return;
ProgressStart("Find edges");
EdgeList Edges;
FindGlobalEdges(Edges, MaxImageCount);
ProgressDone();
Progress("%d edges", (int) Edges.size());
ProgressStart("Find families");
FamList Fams;
FindConnectedComponents(Edges, Fams, g_paramMinFamSize);
AssignFamsToPiles(Fams);
ProgressDone();
Progress("%d families", (int) Fams.size());
ProgressStart("Find superfamilies");
EdgeList SuperEdges;
FindSuperFamEdges(Fams, SuperEdges);
FamList SuperFams;
FindConnectedComponents(SuperEdges, SuperFams, 1);
FindSingletonSuperFams(Fams, SuperFams);
AssignSuperFamsToPiles(Fams, SuperFams);
ProgressDone();
Progress("%d superfamilies", (int) SuperFams.size());
ProgressStart("Write TRS output file");
WriteTRSFile(OutputFileName, Piles, PileCount);
ProgressDone();
}
static void CreatePiles(const HitData *Hits, int HitCount,
PILE_INDEX_TYPE *PileIndexes)
{
Piles = all(PileData, PileCount);
zero(Piles, PileData, PileCount);
for (int i = 0; i < PileCount; ++i)
{
Piles[i].FamIndex = -1;
Piles[i].SuperFamIndex = -1;
Piles[i].Rev = -1;
}
// Count images in stack
ProgressStart("Create stacks: count images");
for (int HitIndex = 0; HitIndex < HitCount; ++HitIndex)
{
const HitData &Hit = Hits[HitIndex];
int Pos = Hit.QueryFrom/CHUNK_LENGTH;
PILE_INDEX_TYPE PileIndex = PileIndexes[Pos];
assert(PileIndex == PileIndexes[Hit.QueryTo/CHUNK_LENGTH]);
assert(PileIndex >= 0 && PileIndex < PileCount);
++(Piles[PileIndex].ImageCount);
Pos = Hit.TargetFrom/CHUNK_LENGTH;
PileIndex = PileIndexes[Pos];
assert(PileIndex >= 0 && PileIndex < PileCount);
assert(PileIndex == PileIndexes[Hit.TargetTo/CHUNK_LENGTH]);
++(Piles[PileIndex].ImageCount);
}
ProgressDone();
// Allocate memory for image list
int TotalImageCount = 0;
ProgressStart("Create stacks: allocate image memory");
for (int PileIndex = 0; PileIndex < PileCount; ++PileIndex)
{
PileData &Pile = Piles[PileIndex];
const int ImageCount = Pile.ImageCount;
TotalImageCount += ImageCount;
assert(ImageCount > 0);
Pile.Images = all(PileImageData, ImageCount);
}
ProgressDone();
// Build image list
ProgressStart("Create stacks: build image list");
for (int PileIndex = 0; PileIndex < PileCount; ++PileIndex)
{
PileData &Pile = Piles[PileIndex];
Pile.ImageCount = 0;
Pile.From = -1;
Pile.To = -1;
}
for (int HitIndex = 0; HitIndex < HitCount; ++HitIndex)
{
const HitData &Hit = Hits[HitIndex];
const bool Rev = Hit.Rev;
const int Length1 = Hit.QueryTo - Hit.QueryFrom;
const int Length2 = Hit.TargetTo - Hit.TargetFrom;
const int From1 = Hit.QueryFrom;
const int From2 = Hit.TargetFrom;
const int To1 = Hit.QueryTo;
const int To2 = Hit.TargetTo;
const int Pos1 = From1/CHUNK_LENGTH;
const int Pos2 = From2/CHUNK_LENGTH;
PILE_INDEX_TYPE PileIndex1 = PileIndexes[Pos1];
PILE_INDEX_TYPE PileIndex2 = PileIndexes[Pos2];
assert(PileIndex1 == PileIndexes[(From1 + Length1 - 1)/CHUNK_LENGTH]);
assert(PileIndex1 >= 0 && PileIndex1 < PileCount);
assert(PileIndex2 == PileIndexes[(From2 + Length2 - 1)/CHUNK_LENGTH]);
assert(PileIndex2 >= 0 && PileIndex2 < PileCount);
PileData &Pile1 = Piles[PileIndex1];
PileImageData &Image1 = Pile1.Images[Pile1.ImageCount++];
Image1.SILength = Length2;
Image1.SIPile = PileIndex2;
Image1.SIRev = Rev;
PileData &Pile2 = Piles[PileIndex2];
PileImageData &Image2 = Pile2.Images[Pile2.ImageCount++];
Image2.SILength = Length1;
Image2.SIPile = PileIndex1;
Image2.SIRev = Rev;
if (Pile1.From == -1 || From1 < Pile1.From)
Pile1.From = From1;
if (Pile1.To == -1 || To1 > Pile1.To)
Pile1.To = To1;
if (Pile2.From == -1 || From2 < Pile2.From)
Pile2.From = From2;
if (Pile2.To == -1 || To2 > Pile2.To)
Pile2.To = To2;
if (Pile1.ImageCount > MaxImageCount)
MaxImageCount = Pile1.ImageCount;
if (Pile2.ImageCount > MaxImageCount)
MaxImageCount = Pile2.ImageCount;
}
ProgressDone();
}
LRESULT CPIMDIFrameWndEx::OnProgressDone(WPARAM wParam, LPARAM lParam)
{
ProgressDone();
return 0;
}