bool IsConstructedClusterNotErroneous(size_t constr_cluster_ind) {
vector<string> read_names = GetReadNamesFromCluster(constr_cluster_ind,
constructed_clusters_, constructed_reads_);
set<size_t> orig_clusters_inds;
for(auto it = read_names.begin(); it != read_names.end(); it++)
orig_clusters_inds.insert(original_clusters_.GetCluster(*it));
return orig_clusters_inds.size() == 1;
}
bool IsClusterNotMerged(size_t orig_cluster_ind) {
vector<string> read_names = GetReadNamesFromCluster(orig_cluster_ind, original_clusters_, original_reads_);
set<size_t> constr_cluster_inds;
for(auto it = read_names.begin(); it != read_names.end(); it++)
if(constructed_clusters_.ReadExists(*it))
constr_cluster_inds.insert(constructed_clusters_.GetCluster(*it));
return constr_cluster_inds.size() > 1;
}
size_t GetOriginalSuperCluster(size_t constr_cluster_ind) {
vector<string> read_names = GetReadNamesFromCluster(constr_cluster_ind,
constructed_clusters_, constructed_reads_);
set<size_t> orig_clusters_inds;
for(auto it = read_names.begin(); it != read_names.end(); it++)
orig_clusters_inds.insert(original_clusters_.GetCluster(*it));
assert(orig_clusters_inds.size() == 1);
return *(orig_clusters_inds.begin());
}
void ComputeConstructedSingletons() {
for(auto it = constructed_clusters_.clusters_begin(); it != constructed_clusters_.clusters_end(); it++)
if(constructed_clusters_.IsClusterSingleton(it->first)) {
metrics_.constructed_metrics.num_singletons++;
auto read_name = constructed_reads_[(it->second)[0]];
if(original_clusters_.IsClusterSingleton(original_clusters_.GetCluster(read_name)))
metrics_.constructed_metrics.num_corr_singletons++;
}
metrics_.correct_singletons = metrics_.constructed_metrics.num_corr_singletons;
}
bool IsClusterBigAndSingletons(size_t orig_cluster_ind) {
assert(IsClusterNotMerged(orig_cluster_ind));
vector<string> read_names = GetReadNamesFromCluster(orig_cluster_ind, original_clusters_, original_reads_);
set<size_t> not_trivial_clusters;
for(auto it = read_names.begin(); it != read_names.end(); it++) {
if(constructed_clusters_.GetClusterSizeByReadName(*it) > 1) {
not_trivial_clusters.insert(constructed_clusters_.GetCluster(*it));
}
}
return not_trivial_clusters.size() == 1;
}
double ComputeNotMergedFillin(size_t orig_cluster_ind) {
vector<string> read_names = GetReadNamesFromCluster(orig_cluster_ind, original_clusters_, original_reads_);
map<size_t, vector<size_t> > constr_subclusters;
for(auto it = read_names.begin(); it != read_names.end(); it++) {
size_t cluster_ind = constructed_clusters_.GetCluster(*it);
constr_subclusters[constructed_clusters_.ClusterSize(cluster_ind)].push_back(cluster_ind);
}
for(auto it = constr_subclusters.rbegin(); it != constr_subclusters.rend(); it++) {
auto clusters = it->second;
for(auto cluster_ind = clusters.begin(); cluster_ind != clusters.end(); cluster_ind++)
if(IsConstructedClusterNotErroneous(*cluster_ind)) {
return double(constructed_clusters_.ClusterSize(*cluster_ind)) / original_clusters_.ClusterSize(orig_cluster_ind);
}
}
return 0;
}
