int GetWeightedPageRankMP1(const PNEANet Graph, TIntFltH& PRankH, const TStr& Attr, const double& C, const double& Eps, const int& MaxIter) {
if (!Graph->IsFltAttrE(Attr)) return -1;
TFltV Weights = Graph->GetFltAttrVecE(Attr);
int mxid = Graph->GetMxNId();
TFltV OutWeights(mxid);
Graph->GetWeightOutEdgesV(OutWeights, Weights);
/*for (TNEANet::TNodeI NI = Graph->BegNI(); NI < Graph->EndNI(); NI++) {
OutWeights[NI.GetId()] = Graph->GetWeightOutEdges(NI, Attr);
}*/
/*TIntFltH Weights;
for (TNEANet::TNodeI NI = Graph->BegNI(); NI < Graph->EndNI(); NI++) {
Weights.AddDat(NI.GetId(), Graph->GetWeightOutEdges(NI, Attr));
}*/
const int NNodes = Graph->GetNodes();
TVec<TNEANet::TNodeI> NV;
//const double OneOver = 1.0/double(NNodes);
PRankH.Gen(NNodes);
for (TNEANet::TNodeI NI = Graph->BegNI(); NI < Graph->EndNI(); NI++) {
NV.Add(NI);
PRankH.AddDat(NI.GetId(), 1.0/NNodes);
//IAssert(NI.GetId() == PRankH.GetKey(PRankH.Len()-1));
}
TFltV TmpV(NNodes);
for (int iter = 0; iter < MaxIter; iter++) {
#pragma omp parallel for schedule(dynamic,10000)
for (int j = 0; j < NNodes; j++) {
TNEANet::TNodeI NI = NV[j];
TmpV[j] = 0;
for (int e = 0; e < NI.GetInDeg(); e++) {
const int InNId = NI.GetInNId(e);
const TFlt OutWeight = OutWeights[InNId];
int EId = Graph->GetEId(InNId, NI.GetId());
const TFlt Weight = Weights[Graph->GetFltKeyIdE(EId)];
if (OutWeight > 0) {
TmpV[j] += PRankH.GetDat(InNId) * Weight / OutWeight; }
}
TmpV[j] = C*TmpV[j]; // Berkhin (the correct way of doing it)
//TmpV[j] = C*TmpV[j] + (1.0-C)*OneOver; // iGraph
}
double diff=0, sum=0, NewVal;
#pragma omp parallel for reduction(+:sum) schedule(dynamic,10000)
for (int i = 0; i < TmpV.Len(); i++) { sum += TmpV[i]; }
const double Leaked = (1.0-sum) / double(NNodes);
#pragma omp parallel for reduction(+:diff) schedule(dynamic,10000)
for (int i = 0; i < PRankH.Len(); i++) { // re-instert leaked PageRank
NewVal = TmpV[i] + Leaked; // Berkhin
//NewVal = TmpV[i] / sum; // iGraph
diff += fabs(NewVal-PRankH[i]);
PRankH[i] = NewVal;
}
if (diff < Eps) { break; }
}
return 0;
}
void DoMuscle()
{
SetOutputFileName(g_pstrOutFileName.get());
SetInputFileName(g_pstrInFileName.get());
SetMaxIters(g_uMaxIters.get());
SetSeqWeightMethod(g_SeqWeight1.get());
TextFile fileIn(g_pstrInFileName.get());
SeqVect v;
v.FromFASTAFile(fileIn);
const unsigned uSeqCount = v.Length();
if (0 == uSeqCount)
Quit("No sequences in input file");
ALPHA Alpha = ALPHA_Undefined;
switch (g_SeqType.get())
{
case SEQTYPE_Auto:
Alpha = v.GuessAlpha();
break;
case SEQTYPE_Protein:
Alpha = ALPHA_Amino;
break;
case SEQTYPE_DNA:
Alpha = ALPHA_DNA;
break;
case SEQTYPE_RNA:
Alpha = ALPHA_RNA;
break;
default:
Quit("Invalid seq type");
}
SetAlpha(Alpha);
v.FixAlpha();
//
// AED 21/12/06: Moved matrix loading code inside the PP param function so it gets called for all alignment types
//
SetPPScore();
unsigned uMaxL = 0;
unsigned uTotL = 0;
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
{
unsigned L = v.GetSeq(uSeqIndex).Length();
uTotL += L;
if (L > uMaxL)
uMaxL = L;
}
SetIter(1);
g_bDiags.get() = g_bDiags1.get();
SetSeqStats(uSeqCount, uMaxL, uTotL/uSeqCount);
SetMuscleSeqVect(v);
MSA::SetIdCount(uSeqCount);
// Initialize sequence ids.
// From this point on, ids must somehow propogate from here.
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
v.SetSeqId(uSeqIndex, uSeqIndex);
if (0 == uSeqCount)
Quit("Input file '%s' has no sequences", g_pstrInFileName.get());
if (1 == uSeqCount)
{
TextFile fileOut(g_pstrOutFileName.get(), true);
v.ToFile(fileOut);
return;
}
if (uSeqCount > 1)
MHackStart(v);
// First iteration
Tree GuideTree;
if (0 != g_pstrUseTreeFileName.get())
{
// Discourage users...
if (!g_bUseTreeNoWarn.get())
fprintf(stderr, g_strUseTreeWarning);
// Read tree from file
TextFile TreeFile(g_pstrUseTreeFileName.get());
GuideTree.FromFile(TreeFile);
// Make sure tree is rooted
if (!GuideTree.IsRooted())
Quit("User tree must be rooted");
if (GuideTree.GetLeafCount() != uSeqCount)
Quit("User tree does not match input sequences");
const unsigned uNodeCount = GuideTree.GetNodeCount();
for (unsigned uNodeIndex = 0; uNodeIndex < uNodeCount; ++uNodeIndex)
{
if (!GuideTree.IsLeaf(uNodeIndex))
continue;
const char *LeafName = GuideTree.GetLeafName(uNodeIndex);
unsigned uSeqIndex;
bool SeqFound = v.FindName(LeafName, &uSeqIndex);
if (!SeqFound)
Quit("Label %s in tree does not match sequences", LeafName);
unsigned uId = v.GetSeqIdFromName(LeafName);
GuideTree.SetLeafId(uNodeIndex, uId);
}
}
else
TreeFromSeqVect(v, GuideTree, g_Cluster1.get(), g_Distance1.get(), g_Root1.get(),
g_pstrDistMxFileName1.get());
const char *Tree1 = ValueOpt("Tree1");
if (0 != Tree1)
{
TextFile f(Tree1, true);
GuideTree.ToFile(f);
if (g_bClusterOnly.get())
return;
}
SetMuscleTree(GuideTree);
ValidateMuscleIds(GuideTree);
MSA msa;
ProgNode *ProgNodes = 0;
if (g_bLow.get())
ProgNodes = ProgressiveAlignE(v, GuideTree, msa);
else
ProgressiveAlign(v, GuideTree, msa);
SetCurrentAlignment(msa);
if (0 != g_pstrComputeWeightsFileName.get())
{
extern void OutWeights(const char *FileName, const MSA &msa);
SetMSAWeightsMuscle(msa);
OutWeights(g_pstrComputeWeightsFileName.get(), msa);
return;
}
ValidateMuscleIds(msa);
if (1 == g_uMaxIters.get() || 2 == uSeqCount)
{
//TextFile fileOut(g_pstrOutFileName.get(), true);
//MHackEnd(msa);
//msa.ToFile(fileOut);
MuscleOutput(msa);
return;
}
if (0 == g_pstrUseTreeFileName.get())
{
g_bDiags.get() = g_bDiags2.get();
SetIter(2);
if (g_bLow.get())
{
if (0 != g_uMaxTreeRefineIters.get())
RefineTreeE(msa, v, GuideTree, ProgNodes);
}
else
RefineTree(msa, GuideTree);
const char *Tree2 = ValueOpt("Tree2");
if (0 != Tree2)
{
TextFile f(Tree2, true);
GuideTree.ToFile(f);
}
}
SetSeqWeightMethod(g_SeqWeight2.get());
SetMuscleTree(GuideTree);
if (g_bAnchors.get())
RefineVert(msa, GuideTree, g_uMaxIters.get() - 2);
else
RefineHoriz(msa, GuideTree, g_uMaxIters.get() - 2, false, false);
#if 0
// Refining by subfamilies is disabled as it didn't give better
// results. I tried doing this before and after RefineHoriz.
// Should get back to this as it seems like this should work.
RefineSubfams(msa, GuideTree, g_uMaxIters.get() - 2);
#endif
ValidateMuscleIds(msa);
ValidateMuscleIds(GuideTree);
//TextFile fileOut(g_pstrOutFileName.get(), true);
//MHackEnd(msa);
//msa.ToFile(fileOut);
MuscleOutput(msa);
}
void DoMuscle(CompositeVect*CVLocation)
{
SetOutputFileName(g_pstrOutFileName);
SetInputFileName(g_pstrInFileName);
SetMaxIters(g_uMaxIters);
SetSeqWeightMethod(g_SeqWeight1);
TextFile fileIn(g_pstrInFileName);
SeqVect v;
v.FromFASTAFile(fileIn);
const unsigned uSeqCount = v.Length();
if (0 == uSeqCount)
Quit("No sequences in input file");
ALPHA Alpha = ALPHA_Undefined;
switch (g_SeqType)
{
case SEQTYPE_Auto:
Alpha = v.GuessAlpha();
break;
case SEQTYPE_Protein:
Alpha = ALPHA_Amino;
break;
case SEQTYPE_DNA:
Alpha = ALPHA_DNA;
break;
case SEQTYPE_RNA:
Alpha = ALPHA_RNA;
break;
default:
Quit("Invalid seq type");
}
SetAlpha(Alpha);
v.FixAlpha();
PTR_SCOREMATRIX UserMatrix = 0;
if (0 != g_pstrMatrixFileName)
{
const char *FileName = g_pstrMatrixFileName;
const char *Path = getenv("MUSCLE_MXPATH");
if (Path != 0)
{
size_t n = strlen(Path) + 1 + strlen(FileName) + 1;
char *NewFileName = new char[n];
sprintf(NewFileName, "%s/%s", Path, FileName);
FileName = NewFileName;
}
TextFile File(FileName);
UserMatrix = ReadMx(File);
g_Alpha = ALPHA_Amino;
g_PPScore = PPSCORE_SP;
}
SetPPScore();
if (0 != UserMatrix)
g_ptrScoreMatrix = UserMatrix;
unsigned uMaxL = 0;
unsigned uTotL = 0;
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
{
unsigned L = v.GetSeq(uSeqIndex).Length();
uTotL += L;
if (L > uMaxL)
uMaxL = L;
}
SetIter(1);
g_bDiags = g_bDiags1;
SetSeqStats(uSeqCount, uMaxL, uTotL/uSeqCount);
SetMuscleSeqVect(v);
MSA::SetIdCount(uSeqCount);
// Initialize sequence ids.
// From this point on, ids must somehow propogate from here.
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
v.SetSeqId(uSeqIndex, uSeqIndex);
if (0 == uSeqCount)
Quit("Input file '%s' has no sequences", g_pstrInFileName);
if (1 == uSeqCount)
{
TextFile fileOut(g_pstrOutFileName, true);
v.ToFile(fileOut);
return;
}
if (uSeqCount > 1)
MHackStart(v);
// First iteration
Tree GuideTree;
if (0 != g_pstrUseTreeFileName)
{
// Discourage users...
if (!g_bUseTreeNoWarn)
fprintf(stderr, "%s", g_strUseTreeWarning);
// Read tree from file
TextFile TreeFile(g_pstrUseTreeFileName);
GuideTree.FromFile(TreeFile);
// Make sure tree is rooted
if (!GuideTree.IsRooted())
Quit("User tree must be rooted");
if (GuideTree.GetLeafCount() != uSeqCount)
Quit("User tree does not match input sequences");
const unsigned uNodeCount = GuideTree.GetNodeCount();
for (unsigned uNodeIndex = 0; uNodeIndex < uNodeCount; ++uNodeIndex)
{
if (!GuideTree.IsLeaf(uNodeIndex))
continue;
const char *LeafName = GuideTree.GetLeafName(uNodeIndex);
unsigned uSeqIndex;
bool SeqFound = v.FindName(LeafName, &uSeqIndex);
if (!SeqFound)
Quit("Label %s in tree does not match sequences", LeafName);
unsigned uId = v.GetSeqIdFromName(LeafName);
GuideTree.SetLeafId(uNodeIndex, uId);
}
}
else
TreeFromSeqVect(v, GuideTree, g_Cluster1, g_Distance1, g_Root1,
g_pstrDistMxFileName1);
const char *Tree1 = ValueOpt("Tree1");
if (0 != Tree1)
{
TextFile f(Tree1, true);
GuideTree.ToFile(f);
if (g_bClusterOnly)
return;
}
SetMuscleTree(GuideTree);
ValidateMuscleIds(GuideTree);
MSA msa;
msa.SetCompositeVector(CVLocation);
ProgNode *ProgNodes = 0;
if (g_bLow)
ProgNodes = ProgressiveAlignE(v, GuideTree, msa);
else
ProgressiveAlign(v, GuideTree, msa);
SetCurrentAlignment(msa);
if (0 != g_pstrComputeWeightsFileName)
{
extern void OutWeights(const char *FileName, const MSA &msa);
SetMSAWeightsMuscle(msa);
OutWeights(g_pstrComputeWeightsFileName, msa);
return;
}
ValidateMuscleIds(msa);
if (1 == g_uMaxIters || 2 == uSeqCount)
{
//TextFile fileOut(g_pstrOutFileName, true);
//MHackEnd(msa);
//msa.ToFile(fileOut);
MuscleOutput(msa);
return;
}
if (0 == g_pstrUseTreeFileName)
{
g_bDiags = g_bDiags2;
SetIter(2);
if (g_bLow)
{
if (0 != g_uMaxTreeRefineIters)
RefineTreeE(msa, v, GuideTree, ProgNodes);
}
else
RefineTree(msa, GuideTree);
const char *Tree2 = ValueOpt("Tree2");
if (0 != Tree2)
{
TextFile f(Tree2, true);
GuideTree.ToFile(f);
}
}
SetSeqWeightMethod(g_SeqWeight2);
SetMuscleTree(GuideTree);
if (g_bAnchors)
RefineVert(msa, GuideTree, g_uMaxIters - 2);
else
RefineHoriz(msa, GuideTree, g_uMaxIters - 2, false, false);
#if 0
// Refining by subfamilies is disabled as it didn't give better
// results. I tried doing this before and after RefineHoriz.
// Should get back to this as it seems like this should work.
RefineSubfams(msa, GuideTree, g_uMaxIters - 2);
#endif
ValidateMuscleIds(msa);
ValidateMuscleIds(GuideTree);
//TextFile fileOut(g_pstrOutFileName, true);
//MHackEnd(msa);
//msa.ToFile(fileOut);
MuscleOutput(msa);
}
int GetWeightedPageRankMP2(const PNEANet Graph, TIntFltH& PRankH, const TStr& Attr, const double& C, const double& Eps, const int& MaxIter) {
if (!Graph->IsFltAttrE(Attr)) return -1;
const int NNodes = Graph->GetNodes();
TVec<TNEANet::TNodeI> NV;
//const double OneOver = 1.0/double(NNodes);
PRankH.Gen(NNodes);
int MxId;
for (TNEANet::TNodeI NI = Graph->BegNI(); NI < Graph->EndNI(); NI++) {
NV.Add(NI);
PRankH.AddDat(NI.GetId(), 1.0/NNodes);
int Id = NI.GetId();
if (Id > MxId) {
MxId = Id;
}
}
TFltV PRankV(MxId+1);
TFltV OutWeights(MxId+1);
TFltV Weights = Graph->GetFltAttrVecE(Attr);
#pragma omp parallel for schedule(dynamic,10000)
for (int j = 0; j < NNodes; j++) {
TNEANet::TNodeI NI = NV[j];
int Id = NI.GetId();
OutWeights[Id] = Graph->GetWeightOutEdges(NI, Attr);
PRankV[Id] = 1/NNodes;
}
TFltV TmpV(NNodes);
for (int iter = 0; iter < MaxIter; iter++) {
#pragma omp parallel for schedule(dynamic,10000)
for (int j = 0; j < NNodes; j++) {
TNEANet::TNodeI NI = NV[j];
TFlt Tmp = 0;
for (int e = 0; e < NI.GetInDeg(); e++) {
const int InNId = NI.GetInNId(e);
const TFlt OutWeight = OutWeights[InNId];
int EId = Graph->GetEId(InNId, NI.GetId());
const TFlt Weight = Weights[Graph->GetFltKeyIdE(EId)];
if (OutWeight > 0) {
Tmp += PRankH.GetDat(InNId) * Weight / OutWeight;
}
}
TmpV[j] = C*Tmp; // Berkhin (the correct way of doing it)
//TmpV[j] = C*TmpV[j] + (1.0-C)*OneOver; // iGraph
}
double sum = 0;
#pragma omp parallel for reduction(+:sum) schedule(dynamic,10000)
for (int i = 0; i < TmpV.Len(); i++) { sum += TmpV[i]; }
const double Leaked = (1.0-sum) / double(NNodes);
double diff = 0;
#pragma omp parallel for reduction(+:diff) schedule(dynamic,10000)
for (int i = 0; i < NNodes; i++) {
TNEANet::TNodeI NI = NV[i];
double NewVal = TmpV[i] + Leaked; // Berkhin
//NewVal = TmpV[i] / sum; // iGraph
int Id = NI.GetId();
diff += fabs(NewVal-PRankV[Id]);
PRankV[Id] = NewVal;
}
if (diff < Eps) { break; }
}
#pragma omp parallel for schedule(dynamic,10000)
for (int i = 0; i < NNodes; i++) {
TNEANet::TNodeI NI = NV[i];
PRankH[i] = PRankV[NI.GetId()];
}
return 0;
}
