// Perecent group identity of a pair of sequences.
// Positions with one or both gapped are ignored.
double MSA::GetPctGroupIdentityPair(unsigned uSeqIndex1,
unsigned uSeqIndex2) const
{
const unsigned uColCount = GetColCount();
unsigned uPosCount = 0;
unsigned uSameCount = 0;
for (unsigned uColIndex = 0; uColIndex < uColCount; ++uColIndex)
{
if (IsGap(uSeqIndex1, uColIndex))
continue;
if (IsGap(uSeqIndex2, uColIndex))
continue;
if (IsWildcard(uSeqIndex1, uColIndex))
continue;
if (IsWildcard(uSeqIndex2, uColIndex))
continue;
const unsigned uLetter1 = GetLetter(uSeqIndex1, uColIndex);
const unsigned uLetter2 = GetLetter(uSeqIndex2, uColIndex);
const unsigned uGroup1 = ResidueGroup[uLetter1];
const unsigned uGroup2 = ResidueGroup[uLetter2];
if (uGroup1 == uGroup2)
++uSameCount;
++uPosCount;
}
if (0 == uPosCount)
return 0;
return (double) uSameCount / (double) uPosCount;
}
double MSA::GetCons(unsigned uColIndex) const
{
unsigned Counts[MAX_ALPHA];
for (unsigned uLetter = 0; uLetter < g_AlphaSize.get(); ++uLetter)
Counts[uLetter] = 0;
unsigned uMaxCount = 0;
for (unsigned uSeqIndex = 0; uSeqIndex < GetSeqCount(); ++uSeqIndex)
{
if (IsGap(uSeqIndex, uColIndex))
continue;
char c = GetChar(uSeqIndex, uColIndex);
c = toupper(c);
if ('X' == c || 'B' == c || 'Z' == c)
continue;
unsigned uLetter = GetLetter(uSeqIndex, uColIndex);
unsigned uCount = Counts[uLetter] + 1;
if (uCount > uMaxCount)
uMaxCount = uCount;
Counts[uLetter] = uCount;
}
// Cons is undefined for all-gap column
if (0 == uMaxCount)
{
// assert(false);
return 1;
}
double dCons = (double) uMaxCount / (double) GetSeqCount();
assert(dCons > 0 && dCons <= 1);
return dCons;
}
void MSA::GetFractionalWeightedCounts(unsigned uColIndex, bool bNormalize,
FCOUNT fcCounts[], FCOUNT *ptrfcGapStart, FCOUNT *ptrfcGapEnd,
FCOUNT *ptrfcGapExtend, FCOUNT *ptrfOcc,
FCOUNT *ptrfcLL, FCOUNT *ptrfcLG, FCOUNT *ptrfcGL, FCOUNT *ptrfcGG) const
{
const unsigned uSeqCount = GetSeqCount();
const unsigned uColCount = GetColCount();
const char* seqName;
memset(fcCounts, 0, g_AlphaSize*sizeof(FCOUNT));
WEIGHT wTotal = 0;
FCOUNT fGap = 0;
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
{
const WEIGHT w = GetSeqWeight(uSeqIndex);
if (IsGap(uSeqIndex, uColIndex))
{
fGap += w;
continue;
}
else if (IsWildcard(uSeqIndex, uColIndex))
{
const unsigned uLetter = GetLetterEx(uSeqIndex, uColIndex);
switch (g_Alpha)
{
case ALPHA_Amino:
switch (uLetter)
{
case AX_B: // D or N
fcCounts[AX_D] += w/2;
fcCounts[AX_N] += w/2;
break;
case AX_Z: // E or Q
fcCounts[AX_E] += w/2;
fcCounts[AX_Q] += w/2;
break;
default: // any
{
const FCOUNT f = w/20;
for (unsigned uLetter = 0; uLetter < 20; ++uLetter)
fcCounts[uLetter] += f;
break;
}
}
break;
case ALPHA_DNA:
case ALPHA_RNA:
switch (uLetter)
{
case AX_R: // G or A
fcCounts[NX_G] += w/2;
fcCounts[NX_A] += w/2;
break;
case AX_Y: // C or T/U
fcCounts[NX_C] += w/2;
fcCounts[NX_T] += w/2;
break;
default: // any
const FCOUNT f = w/20;
for (unsigned uLetter = 0; uLetter < 4; ++uLetter)
fcCounts[uLetter] += f;
break;
}
break;
default:
Quit("Alphabet %d not supported", g_Alpha);
}
continue;
}
unsigned uLetter = GetLetter(uSeqIndex, uColIndex);
//BEGIN MODIFICATIONS TO MUSCLE
int original=0;
for(unsigned i=0; i<uColIndex; i++){
if (i >= this->GetColCount()){break;}
++original;
char c = GetChar(uSeqIndex, i);
if(c== '-'){
original--;
}
}
seqName = this->GetSeqName(uSeqIndex);
int compositeVectPosition;
compositeVectPosition = atoi(seqName);
CompositeVect CV = *CVLocation;
Composite* CVL = CV[compositeVectPosition];
Composite C = *CVL;
for(int j=0; j<21; j++){
fcCounts[j] = w*C[original][j];
wTotal = w*C[original][j];
}
//ORIGINAL MUSLCE LINE WAS:
//fcCounts[uLetter] += w;
//wTotal += w;
//END MODIFICATIONS TO MUSCLE
}
*ptrfOcc = (float) (1.0 - fGap);
if (bNormalize && wTotal > 0)
{
if (wTotal > 1.001)
Quit("wTotal=%g\n", wTotal);
for (unsigned uLetter = 0; uLetter < g_AlphaSize; ++uLetter)
fcCounts[uLetter] /= wTotal;
// AssertNormalized(fcCounts);
}
FCOUNT fcStartCount = 0;
if (uColIndex == 0)
{
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex))
fcStartCount += GetSeqWeight(uSeqIndex);
}
else
{
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex) && !IsGap(uSeqIndex, uColIndex - 1))
fcStartCount += GetSeqWeight(uSeqIndex);
}
FCOUNT fcEndCount = 0;
if (uColCount - 1 == uColIndex)
{
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex))
fcEndCount += GetSeqWeight(uSeqIndex);
}
else
{
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex) && !IsGap(uSeqIndex, uColIndex + 1))
fcEndCount += GetSeqWeight(uSeqIndex);
}
FCOUNT LL = 0;
FCOUNT LG = 0;
FCOUNT GL = 0;
FCOUNT GG = 0;
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
{
WEIGHT w = GetSeqWeight(uSeqIndex);
bool bLetterHere = !IsGap(uSeqIndex, uColIndex);
bool bLetterPrev = (uColIndex == 0 || !IsGap(uSeqIndex, uColIndex - 1));
if (bLetterHere)
{
if (bLetterPrev)
LL += w;
else
GL += w;
}
else
{
if (bLetterPrev)
LG += w;
else
GG += w;
}
}
FCOUNT fcExtendCount = 0;
if (uColIndex > 0 && uColIndex < GetColCount() - 1)
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex) && IsGap(uSeqIndex, uColIndex - 1) &&
IsGap(uSeqIndex, uColIndex + 1))
fcExtendCount += GetSeqWeight(uSeqIndex);
*ptrfcLL = LL;
*ptrfcLG = LG;
*ptrfcGL = GL;
*ptrfcGG = GG;
*ptrfcGapStart = fcStartCount;
*ptrfcGapEnd = fcEndCount;
*ptrfcGapExtend = fcExtendCount;
}
void MSA::GetFractionalWeightedCounts(unsigned uColIndex, bool bNormalize,
FCOUNT fcCounts[], FCOUNT *ptrfcGapStart, FCOUNT *ptrfcGapEnd,
FCOUNT *ptrfcGapExtend, FCOUNT *ptrfOcc,
FCOUNT *ptrfcLL, FCOUNT *ptrfcLG, FCOUNT *ptrfcGL, FCOUNT *ptrfcGG) const
{
const unsigned uSeqCount = GetSeqCount();
const unsigned uColCount = GetColCount();
memset(fcCounts, 0, g_AlphaSize*sizeof(FCOUNT));
WEIGHT wTotal = 0;
FCOUNT fGap = 0;
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
{
const WEIGHT w = GetSeqWeight(uSeqIndex);
if (IsGap(uSeqIndex, uColIndex))
{
fGap += w;
continue;
}
else if (IsWildcard(uSeqIndex, uColIndex))
{
const unsigned uLetter = GetLetterEx(uSeqIndex, uColIndex);
switch (g_Alpha)
{
case ALPHA_Amino:
switch (uLetter)
{
case AX_B: // D or N
fcCounts[AX_D] += w/2;
fcCounts[AX_N] += w/2;
break;
case AX_Z: // E or Q
fcCounts[AX_E] += w/2;
fcCounts[AX_Q] += w/2;
break;
default: // any
{
const FCOUNT f = w/20;
for (unsigned uLetter = 0; uLetter < 20; ++uLetter)
fcCounts[uLetter] += f;
break;
}
}
break;
case ALPHA_DNA:
case ALPHA_RNA:
switch (uLetter)
{
case AX_R: // G or A
fcCounts[NX_G] += w/2;
fcCounts[NX_A] += w/2;
break;
case AX_Y: // C or T/U
fcCounts[NX_C] += w/2;
fcCounts[NX_T] += w/2;
break;
default: // any
const FCOUNT f = w/20;
for (unsigned uLetter = 0; uLetter < 4; ++uLetter)
fcCounts[uLetter] += f;
break;
}
break;
default:
Quit("Alphabet %d not supported", g_Alpha);
}
continue;
}
unsigned uLetter = GetLetter(uSeqIndex, uColIndex);
fcCounts[uLetter] += w;
wTotal += w;
}
*ptrfOcc = (float) (1.0 - fGap);
if (bNormalize && wTotal > 0)
{
if (wTotal > 1.001)
Quit("wTotal=%g\n", wTotal);
for (unsigned uLetter = 0; uLetter < g_AlphaSize; ++uLetter)
fcCounts[uLetter] /= wTotal;
// AssertNormalized(fcCounts);
}
FCOUNT fcStartCount = 0;
if (uColIndex == 0)
{
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex))
fcStartCount += GetSeqWeight(uSeqIndex);
}
else
{
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex) && !IsGap(uSeqIndex, uColIndex - 1))
fcStartCount += GetSeqWeight(uSeqIndex);
}
FCOUNT fcEndCount = 0;
if (uColCount - 1 == uColIndex)
{
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex))
fcEndCount += GetSeqWeight(uSeqIndex);
}
else
{
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex) && !IsGap(uSeqIndex, uColIndex + 1))
fcEndCount += GetSeqWeight(uSeqIndex);
}
FCOUNT LL = 0;
FCOUNT LG = 0;
FCOUNT GL = 0;
FCOUNT GG = 0;
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
{
WEIGHT w = GetSeqWeight(uSeqIndex);
bool bLetterHere = !IsGap(uSeqIndex, uColIndex);
bool bLetterPrev = (uColIndex == 0 || !IsGap(uSeqIndex, uColIndex - 1));
if (bLetterHere)
{
if (bLetterPrev)
LL += w;
else
GL += w;
}
else
{
if (bLetterPrev)
LG += w;
else
GG += w;
}
}
FCOUNT fcExtendCount = 0;
if (uColIndex > 0 && uColIndex < GetColCount() - 1)
for (unsigned uSeqIndex = 0; uSeqIndex < uSeqCount; ++uSeqIndex)
if (IsGap(uSeqIndex, uColIndex) && IsGap(uSeqIndex, uColIndex - 1) &&
IsGap(uSeqIndex, uColIndex + 1))
fcExtendCount += GetSeqWeight(uSeqIndex);
*ptrfcLL = LL;
*ptrfcLG = LG;
*ptrfcGL = GL;
*ptrfcGG = GG;
*ptrfcGapStart = fcStartCount;
*ptrfcGapEnd = fcEndCount;
*ptrfcGapExtend = fcExtendCount;
}
