/* 'ReadBuffer::flush' writes the current contents of the buffer to the
designated output file, and then clears the contents to be ready to receive
the next reads. */
void ReadBuffer::flush()
{
// std::cout << "in flush " << std::endl;
#pragma omp parallel for
for (int i=0; i<m_currentsize ; i++ ) {
// std::cout << "before GetCloseEnd " << std::endl;
GetCloseEnd(m_CHROMOSOME, m_rawreads[i]);
// std::cout << "after GetCloseEnd " << std::endl;
if (m_rawreads[i].hasCloseEnd()) {
//if (m_rawreads[i].Name == "@DD7DT8Q1:4:1106:17724:13906#GTACCT/1") {
// std::cout << "m_rawreads[i].hasCloseEnd()" << std::endl;
//}
updateReadAfterCloseEndMapping(m_rawreads[i]);
#pragma omp critical
m_filteredReads.push_back(m_rawreads[i]);
}
else {
//if (m_rawreads[i].Name == "@DD7DT8Q1:4:1106:17724:13906#GTACCT/1") {
// std::cout << "m_rawreads[i] no close end" << std::endl;
//}
#pragma omp critical
m_OneEndMappedReads.push_back(m_rawreads[i]);
}
}
//std::cout << "end of flush " << std::endl;
m_rawreads.clear();
m_currentsize = 0;
//std::cout << "existing flush " << std::endl;
}
/* 'ReadBuffer::flush' writes the current contents of the buffer to the
designated output file, and then clears the contents to be ready to receive
the next reads. */
void ReadBuffer::flush()
{
// std::cout << "in flush " << std::endl;
#pragma omp parallel for
for (int i=0; i<m_currentsize ; i++ ) {
// std::cout << "before GetCloseEnd " << std::endl;
//std::map<std::string, unsigned>::iterator it = g_ReadSeq2Index.find(m_rawreads[i].UnmatchedSeq);
//if (it == g_ReadSeq2Index.end())
{
//if (m_rawreads[i].MapperSplit ) {
// std::cout << "skip close end search" << std::endl;
//}
if (m_rawreads[i].MapperSplit == false)
GetCloseEnd(m_CHROMOSOME, m_rawreads[i]);
if (m_rawreads[i].hasCloseEnd()) {
updateReadAfterCloseEndMapping(m_rawreads[i]);
#pragma omp critical
{
// g_ReadSeq2Index.insert(std::pair<std::string, unsigned> (m_rawreads[i].UnmatchedSeq, m_filteredReads.size()));
m_rawreads[i].SampleName2Number.insert(std::pair <std::string, unsigned> (m_rawreads[i].Tag, 1));
m_filteredReads.push_back(m_rawreads[i]);
}
}
// else {
//if (m_rawreads[i].Name == "@DD7DT8Q1:4:1106:17724:13906#GTACCT/1") {
// std::cout << "m_rawreads[i] no close end" << std::endl;
//}
//#pragma omp critical
//m_OneEndMappedReads.push_back(m_rawreads[i]);
// }
}
//else { // SampleName2Number std::map <std::string, unsigned> SampleName2Number;
// #pragma omp critical
// {
// unsigned ReadIndex = it -> second; // m_filteredReads[ReadIndex]
// std::map <std::string, unsigned>::iterator it_SampleName = m_filteredReads[ReadIndex].SampleName2Number.find(m_rawreads[i].Tag);
//
// if (it_SampleName == m_filteredReads[ReadIndex].SampleName2Number.end()) {
// //std::cout << "adding " << m_rawreads[i].Tag << "\t1" << std::endl;
// m_filteredReads[ReadIndex].SampleName2Number.insert(std::pair <std::string, unsigned> (m_rawreads[i].Tag, 1));
// }
// else {
//
// it_SampleName -> second++;
// //std::cout << "increasing " << m_rawreads[i].Tag << "\t" << it_SampleName -> second << std::endl;
// }
// }
//}
// std::cout << "after GetCloseEnd " << std::endl;
}
//std::cout << "end of flush " << std::endl;
m_rawreads.clear();
m_currentsize = 0;
//std::cout << "existing flush " << std::endl;
}
void CombineReads(const std::string & CurrentChrSeq, const char & Strand, const std::vector <SPLIT_READ> & input_reads, const std::vector <unsigned int> & Index_Of_Useful_Reads, std::vector <SPLIT_READ> & output_reads) {
//std::cout << "start of CombineReads" << std::endl;
std::string Spacer = "";
unsigned Max_ReadLength = 0;
unsigned Max_AssembledLength = 0;
unsigned Min_LeftMostPos = input_reads[Index_Of_Useful_Reads[0]].LeftMostPos;
SPLIT_READ output_one_read;// = input_reads[Index_Of_Useful_Reads[0]];
unsigned Index2Read4Copy = 0;
//unsigned Min_Close_Size = 10000;
for (unsigned ReadIndex = 0; ReadIndex < Index_Of_Useful_Reads.size(); ReadIndex++) {
//if (input_reads[Index_Of_Useful_Reads[ReadIndex]].UP_Close.size() < Min_Close_Size) {
// Min_Close_Size = input_reads[Index_Of_Useful_Reads[ReadIndex]].UP_Close.size();
// Index2Read4Copy = ReadIndex; // input_reads[Index_Of_Useful_Reads[Index2Read4Copy]]
//}
//std::cout << Strand << " " << input_reads[Index_Of_Useful_Reads[ReadIndex]].UP_Close.size() << std::endl;
//std::cout << input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos << " " << input_reads[Index_Of_Useful_Reads[ReadIndex]].UnmatchedSeq << std::endl;
if (input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos < (int)Min_LeftMostPos)
Min_LeftMostPos = input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos;
if (input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength() > (short)Max_ReadLength)
Max_ReadLength = input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength();
if (input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos + (unsigned)input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength() > Max_AssembledLength)
Max_AssembledLength = input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos + input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength();
}
Max_AssembledLength = Max_AssembledLength - Min_LeftMostPos;
if ((float)Max_AssembledLength < Max_ReadLength * 1.3) return;
std::cout << "Max_AssembledLength " << Max_AssembledLength << std::endl;
for (unsigned ReadIndex = 0; ReadIndex < Index_Of_Useful_Reads.size(); ReadIndex++) {
Spacer.clear();
if (Strand == '+') {
for (unsigned SpacerIndex = 0; SpacerIndex < Max_AssembledLength + Min_LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength(); SpacerIndex++) Spacer += " ";
//std::cout << Spacer << (input_reads[Index_Of_Useful_Reads[ReadIndex]].UnmatchedSeq) << std::endl;
}
else {
for (unsigned SpacerIndex = 0; SpacerIndex < input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - Min_LeftMostPos; SpacerIndex++)
Spacer += " ";
//std::cout << Spacer << (input_reads[Index_Of_Useful_Reads[ReadIndex]].UnmatchedSeq) << std::endl;
}
}
if (Strand == '+') { // UnmatchedSeq
std::cout << "+ Max_ReadLength " << Max_ReadLength << "\t" << "Min_LeftMostPos " << Min_LeftMostPos
<< "\nref: \n" << ReverseComplement(CurrentChrSeq.substr(Min_LeftMostPos, Max_AssembledLength)) << std::endl;
}
else if (Strand == '-') {
std::cout << "- Max_ReadLength " << Max_ReadLength << "\t" << "Min_LeftMostPos " << Min_LeftMostPos
<< "\nref: \n" << CurrentChrSeq.substr(Min_LeftMostPos, Max_AssembledLength) << std::endl;
}
unsigned Count[5][Max_AssembledLength];
float Ratio[5][Max_AssembledLength];
for (short i = 0; i < 5; i++) {
for (unsigned j = 0; j < Max_AssembledLength; j++) {
Count[i][j] = 0;
Ratio[i][j] = 0.0;
}
}
if (Strand == '+') {
for (unsigned ReadIndex = 0; ReadIndex < Index_Of_Useful_Reads.size(); ReadIndex++) {
//std::cout << std::endl;
for (short BaseIndex = 0; BaseIndex < input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength(); BaseIndex++) {
//std::cout << input_reads[Index_Of_Useful_Reads[ReadIndex]].UnmatchedSeq[BaseIndex];
switch (input_reads[Index_Of_Useful_Reads[ReadIndex]].getUnmatchedSeq()[BaseIndex]) {
case 'A':
Count[0][Max_AssembledLength + Min_LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength() + BaseIndex]++;
break; // 00000000
case 'C':
Count[1][Max_AssembledLength + Min_LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength() + BaseIndex]++;
break; // 00010000
case 'G':
Count[2][Max_AssembledLength + Min_LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength() + BaseIndex]++;
break; // 00100000
case 'T':
Count[3][Max_AssembledLength + Min_LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength() + BaseIndex]++;
break; // 00110000
default:
Count[4][Max_AssembledLength + Min_LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength() + BaseIndex]++;
// 01000000
}
}
}
}
else if (Strand == '-') {
for (unsigned ReadIndex = 0; ReadIndex < Index_Of_Useful_Reads.size(); ReadIndex++) {
//std::cout << std::endl;
for (short BaseIndex = 0; BaseIndex < input_reads[Index_Of_Useful_Reads[ReadIndex]].getReadLength(); BaseIndex++) {
//std::cout << input_reads[Index_Of_Useful_Reads[ReadIndex]].UnmatchedSeq[BaseIndex];
switch (input_reads[Index_Of_Useful_Reads[ReadIndex]].getUnmatchedSeq()[BaseIndex]) {
case 'A':
Count[0][input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - Min_LeftMostPos + BaseIndex]++;
break; // 00000000
case 'C':
Count[1][input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - Min_LeftMostPos + BaseIndex]++;
break; // 00010000
case 'G':
Count[2][input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - Min_LeftMostPos + BaseIndex]++;
break; // 00100000
case 'T':
Count[3][input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - Min_LeftMostPos + BaseIndex]++;
break; // 00110000
default:
Count[4][input_reads[Index_Of_Useful_Reads[ReadIndex]].LeftMostPos - Min_LeftMostPos + BaseIndex]++;
// 01000000
}
}
}
}
float Sum;
for (unsigned PosIndex = 0; PosIndex < Max_AssembledLength; PosIndex++) {
Sum = Count[0][PosIndex] + Count[1][PosIndex] + Count[2][PosIndex] + Count[3][PosIndex] + Count[4][PosIndex];
//std::cout << Count[0][PosIndex] << " " << Count[1][PosIndex] << " " << Count[2][PosIndex] << " " << Count[3][PosIndex] << " " << Count[4][PosIndex] << std::endl;
for (unsigned BaseIndex = 0; BaseIndex < 5; BaseIndex++) {
Ratio[BaseIndex][PosIndex] = Count[BaseIndex][PosIndex] / Sum;
}
}
std::string OutputOneStr = "";
const float RatioCutoff = 0.66;
unsigned Max_Base_Count = 0;
short Max_Base_Count_Index = -1;
for (unsigned PosIndex = 0; PosIndex < Max_AssembledLength; PosIndex++) {
if (Ratio[0][PosIndex] > RatioCutoff) {
OutputOneStr += "A";
continue;
}
if (Ratio[1][PosIndex] > RatioCutoff) {
OutputOneStr += "C";
continue;
}
if (Ratio[2][PosIndex] > RatioCutoff) {
OutputOneStr += "G";
continue;
}
if (Ratio[3][PosIndex] > RatioCutoff) {
OutputOneStr += "T";
continue;
}
for (short BaseIndex = 0; BaseIndex < 4; BaseIndex++) {
if (Count[BaseIndex][PosIndex] > Max_Base_Count && Count[BaseIndex][PosIndex] >= 3)
Max_Base_Count_Index = BaseIndex;
}
//if (Max_Base_Count_Index != -1) {
switch (Max_Base_Count_Index) {
case 0:
OutputOneStr += "A";
break;
case 1:
OutputOneStr += "C";
break;
case 2:
OutputOneStr += "G";
break;
case 3:
OutputOneStr += "T";
break;
case -1:
OutputOneStr += "N";
break;
default:
break;
}
//}
//else OutputOneStr += "N";
}
if (Strand == '+') {
std::cout << "Final merged string +: original\n" << (OutputOneStr) << std::endl;
std::cout << "Final merged string +: convert to ref\n" << ReverseComplement(OutputOneStr) << std::endl;
}
else {
std::cout << "Final merged string -: original\n" << (OutputOneStr) << std::endl;
std::cout << "Final merged string : convert to ref\n" << (OutputOneStr) << std::endl;
}//std::cout << "Final merged string: -\n" << (OutputOneStr) << std::endl;
//std::cout << "here1" << std::endl;
unsigned Count_N = 0;
for (unsigned pos_index = 0; pos_index < OutputOneStr.size(); pos_index++) {
if (OutputOneStr[pos_index] == 'N') Count_N++;
}
if ((float)Count_N < OutputOneStr.size() * 0.05) {
unsigned Min_Close_Size = 10000;
Index2Read4Copy = 0; // if the best one cannot be found due to N or whatever reasons, use the first read as the template for copy.
//std::cout << "Original Index2Read4Copy: " << Index2Read4Copy << std::endl;
for (unsigned ReadIndex = 0; ReadIndex < Index_Of_Useful_Reads.size(); ReadIndex++) {
if (input_reads[Index_Of_Useful_Reads[ReadIndex]].UP_Close.size() < Min_Close_Size && OutputOneStr.find(input_reads[Index_Of_Useful_Reads[ReadIndex]].getUnmatchedSeq()) !=std::string::npos) { // quick fix here: need more work
Min_Close_Size = input_reads[Index_Of_Useful_Reads[ReadIndex]].UP_Close.size();
Index2Read4Copy = ReadIndex; // input_reads[Index_Of_Useful_Reads[Index2Read4Copy]]
//std::cout << "Changed Index2Read4Copy: " << Index2Read4Copy << std::endl;
}
}
//std::cout << "here2" << std::endl;
output_one_read = input_reads[Index_Of_Useful_Reads[Index2Read4Copy]];
//std::cout << "here2a" << std::endl;
output_one_read.setUnmatchedSeq( OutputOneStr );
//update std::map <std::string, int> ReadCountPerSample;
GetReadCountPerSample(input_reads, Index_Of_Useful_Reads, output_one_read);
//std::cout << "here2b" << std::endl;
//std::cout << "Before: " << output_one_read.UP_Close.size() << std::endl;
output_one_read.UP_Close.clear();
//std::cout << "here3" << std::endl;
output_one_read.Thickness = Index_Of_Useful_Reads.size();
//std::cout << "Thickness " << output_one_read.Thickness << std::endl;
GetCloseEnd(CurrentChrSeq, output_one_read);
//std::cout << "After: " << output_one_read.UP_Close.size() << std::endl;
output_reads.push_back(output_one_read);
//std::cout << "here4" << std::endl;
//std::cout << "end of CombineReads" << std::endl;
}
}