unsigned int HypothesisEvaluator::SolveBeginningOfRead(DPTreephaser &working_treephaser, BasecallerRead &master_read,
const vector<string>& Hypotheses, int startFlow) {
//cout << "Hypothesis sequence: " << Hypotheses[0] << endl;
// Solve beginning of maybe clipped read
if (startFlow>0) {
int until_flow = min((startFlow+20), nFlows);
working_treephaser.Solve(master_read, until_flow, 0);
}
/*cout << "Solved prefix of size " << read.sequence.size() << ": ";
for (int i=0; i<read.sequence.size(); i++)
cout << read.sequence[i];
cout << endl;*/
// StartFlow clipped? Get solved HP length at startFlow
unsigned int base = 0;
int flow = 0;
int HPlength = 0;
while (base<master_read.sequence.size()) {
while (flow < treePhaserFlowOrder.num_flows() and treePhaserFlowOrder.nuc_at(flow) != master_read.sequence[base])
flow++;
if (flow > startFlow or flow == treePhaserFlowOrder.num_flows())
break;
if (flow == startFlow)
HPlength++;
base++;
}
// Get HP size at the start of the reference, i.e., Hypotheses[0]
int count = 1;
while (Hypotheses[0][count] == Hypotheses[0][0])
count++;
// Adjust the length of the base prefix and erase extra solved bases
if (HPlength>count)
base -= count;
else
base -= HPlength;
master_read.sequence.erase(master_read.sequence.begin()+base, master_read.sequence.end());
unsigned int prefix_size = master_read.sequence.size();
/*cout << "Shortened prefix to size " << prefix_size << " until startFlow" << startFlow << ": ";
for (int i=0; i<read.sequence.size(); i++)
cout << read.sequence[i];
cout << endl;*/
return(prefix_size);
}
int HypothesisEvaluator::EvaluateOneHypothesis(DPTreephaser &working_treephaser, BasecallerRead ¤t_hypothesis, int applyNormalization) {
int last_incorporating_flow = LastIncorporatingFlow(current_hypothesis);
// Simulate sequence
working_treephaser.Simulate(current_hypothesis, nFlows);
// Adaptively normalize each hypothesis
if (applyNormalization>0) {
int window_size= 50;
int steps = last_incorporating_flow / window_size;
working_treephaser.WindowedNormalize(current_hypothesis, steps, window_size);
}
// Solver simulates beginning of the read and then fills in the remaining clipped bases
working_treephaser.Solve(current_hypothesis, nFlows, last_incorporating_flow);
/*cout << "Solved sequence of length: " << hypothesesReadsVector[r].sequence.size() << " ;nFlows = " << nFlows << endl;
cout << "Total read: ";
for (int i=0; i<hypothesesReadsVector[r].sequence.size(); i++)
cout << hypothesesReadsVector[r].sequence[i];
cout << endl;*/
return(last_incorporating_flow);
}
