/** *************************************************************************** */
public void sortHits() {
// Check for no proteins.
if (total <= 1) return;
// Compare this protein to all other proteins.
AminoWords words = best[0].getWords();
int matches = 0;
int words_total = 0;
for (int i = 1; i < total; i++) {
// Count the number of words common to both sequences.
matches = words.countCommonWords(best[i].getWords());
if (words.getTotal() > 0)
best[i].setPercentFirst((byte) ((matches * 100) / words.getTotal()));
words_total = best[i].getWords().getTotal();
if (words_total < words.getTotal())
best[i].setPercentFirst((byte) ((matches * 100) / words_total));
} // for
// Sort the best hits.
if (total == 2) {
// Find the best matches.
for (int i = 0; i < total; i++) findBest(i);
return;
} // if
Best temp = null;
for (int i = 1; i < total - 1; i++) {
for (int j = i + 1; j < total; j++) {
// Compare by percent words.
if (best[j].getPercentFirst() > best[i].getPercentFirst()) {
temp = best[i];
best[i] = best[j];
best[j] = temp;
} // if
else
// Check for the same percent identity.
if (best[j].getPercentFirst() == best[i].getPercentFirst()) {
if (best[j].getFastaSequence().getLength() > best[i].getFastaSequence().getLength()) {
temp = best[i];
best[i] = best[j];
best[j] = temp;
} // if
} // if
} // for
} // for
// Find the best matches.
for (int i = 0; i < total; i++) findBest(i);
} // method sortHits
/** *************************************************************************** */
private void findBest(int index) {
// Check for the first protein.
if (index <= 0) {
families++; // New protein family
best[index].setFamily(families); // Set the protein family number
super_families++; // New protein super family
best[index].setSuperFamily(super_families); // Set the protein super family number
return;
} // if
// Compare this protein to all other proteins.
int best_count = 0;
int best_match = 0;
int best_percent = 0;
AminoWords words = best[index].getWords();
int length = best[index].getFastaSequence().getLength();
int length2 = length;
for (int i = 0; i < index; i++) {
// Count the number of words common to both sequences.
int count = words.countCommonWords(best[i].getWords());
// Check for a better match.
if (count > best_count) {
best_count = count;
best_match = i;
// best_percent = (count * 100) / best [ i ].getWordsTotal ();
length2 = best[i].getFastaSequence().getLength();
if (length < length2) best_percent = (count * 100) / length;
else best_percent = (count * 100) / length2;
} // if
} // for
// Record the number of common words with the best match.
best[index].setCommonWords(best_count);
best[index].setBestMatch(best_match);
// Check for a family or super-family match.
if (best_count > 9) {
// Check for a family match.
if ((best_count > 49) && (best_percent >= 26)) {
best[index].setFamily(best[best_match].getFamily());
best[index].setSuperFamily(best[best_match].getSuperFamily());
// Check for a better match for the previous sequence.
if (best_count > best[best_match].getCommonWords()) {
best[best_match].setCommonWords(best_count);
best[best_match].setBestMatch(index);
} // if
} // if
else // super family member
{
families++; // New protein family
best[index].setFamily(families); // Set the protein family number
// Set the protein super family number
if (best_count > 19) {
best[index].setSuperFamily(best[best_match].getSuperFamily());
} // if
else {
super_families++; // New protein super family
best[index].setSuperFamily(super_families); // Set the protein super family number
} // else
} // else
} // if
else {
families++; // New protein family
best[index].setFamily(families); // Set the protein family number
super_families++; // New protein super family
best[index].setSuperFamily(super_families); // Set the protein super family number
} // else
} // method findBest
