private void guessHumanReadableName(final CyRow row) {
boolean found = false;
// Special handler for STRING. This is a hack...
if (row.getTable().getColumn(STRING_ATTR_NAME) != null) {
final List<String> stringList = row.getList(STRING_ATTR_NAME, String.class);
if (stringList != null) found = findHumanReadableName(row, stringList, ncbiPattern, true);
}
if (found) return;
// try NCBI
if (row.getTable().getColumn(ENTREZ_GENE_ATTR_NAME) != null) {
final List<String> ncbiList = row.getList(ENTREZ_GENE_ATTR_NAME, String.class);
if (ncbiList != null) found = findHumanReadableName(row, ncbiList, ncbiPattern, true);
}
if (found) return;
// Try Uniprot
List<String> uniprotList = null;
if (row.getTable().getColumn(UNIPROT_ATTR_NAME) != null) {
uniprotList = row.getList(UNIPROT_ATTR_NAME, String.class);
if (uniprotList != null) found = findHumanReadableName(row, uniprotList, exact1Pattern, true);
}
if (found) return;
if (uniprotList != null) found = findHumanReadableName(row, uniprotList, uniprotPattern, false);
if (found) return;
// Unknown
if (row.getTable().getColumn("unknown") != null) {
final List<String> unknownList = row.getList("unknown", String.class);
if (unknownList != null)
found = findHumanReadableName(row, unknownList, uniprotPattern, false);
}
if (found) return;
if (found == false) {
// Give up. Use primary key
row.set(PREDICTED_GENE_NAME, row.get(CyNetwork.NAME, String.class));
}
}
@Test
public void gaReaderTest() throws Exception {
File file = new File("./src/test/resources/" + GA_YEAST);
GeneAssociationReader reader =
new GeneAssociationReader(
"dummy dag", file.toURI().toURL().openStream(), "yeast GA", serviceRegistrar);
System.out.print("Start read: ");
reader.run(tm);
final CyTable[] tables = reader.getTables();
assertNotNull(tables);
assertEquals(1, tables.length);
assertNotNull(tables[0]);
// All 22 Columns + NAME primary key + synonyms
assertEquals(24, tables[0].getColumns().size());
// For yeast test file.
assertEquals(6359, tables[0].getRowCount());
// Check table contents
final CyTable table = tables[0];
final CyRow row1 = table.getRow("S000003319");
assertNotNull(row1);
final List<String> bpList1 = row1.getList("biological process", String.class);
assertNotNull(bpList1);
assertFalse(bpList1.contains("GO:0000287"));
assertTrue(bpList1.contains("GO:0006067"));
assertFalse(bpList1.contains("fake value"));
final String taxName = row1.get(GeneAssociationTag.TAXON.toString(), String.class);
assertNotNull(taxName);
assertEquals("Saccharomyces cerevisiae", taxName);
final List<String> referenceList1 =
row1.getList("biological process DB Reference", String.class);
// assertNotNull(referenceList1);
// assertEquals(2, referenceList1.size());
}
private void processOtherNames(CyRow row, final Map<String, List<String>> accs) {
for (String originalDBName : accs.keySet()) {
final String dbName = validateNamespace(originalDBName);
if (row.getTable().getColumn(dbName) == null)
row.getTable().createListColumn(dbName, String.class, false);
List<String> currentList = row.getList(dbName, String.class);
if (currentList == null) currentList = new ArrayList<String>();
final Set<String> nameSet = new HashSet<String>(currentList);
final List<String> names = accs.get(originalDBName);
nameSet.addAll(names);
row.set(dbName, new ArrayList<String>(nameSet));
}
}
private void mergeRow(String keyName, CyRow sourceRow, CyRow targetRow) {
for (CyColumn column : sourceRow.getTable().getColumns()) {
if (cancelled) return;
String columnName = column.getName();
if (columnName.equals(keyName)) continue;
Class<?> type = column.getType();
if (type.equals(List.class)) {
Class<?> elementType = column.getListElementType();
List<?> list = sourceRow.getList(columnName, elementType);
targetRow.set(columnName, list);
} else {
Object value = sourceRow.get(columnName, type);
targetRow.set(columnName, value);
}
}
}
private void mapEntry(final String[] entries) {
// Set primary key for the table, which is DB Object ID
final String primaryKeyValue = entries[DB_OBJ_ID];
final CyRow row = table.getRow(primaryKeyValue);
row.set(CyNetwork.NAME, primaryKeyValue);
// Check namespace
final String namespace = NAMESPACE_MAP.get(entries[ASPECT]);
for (int i = 0; i < EXPECTED_COL_COUNT; i++) {
final GeneAssociationTag tag = GeneAssociationTag.values()[i];
switch (tag) {
// Evidence code and GO ID should be organized by namespace.
case GO_ID:
String goidString = entries[i];
if (this.termIDList != null) goidString = convertToName(goidString);
List<String> currentList = row.getList(namespace, String.class);
if (currentList == null) currentList = new ArrayList<String>();
if (currentList.contains(goidString) == false) currentList.add(goidString);
row.set(namespace, currentList);
break;
case EVIDENCE:
case DB_REFERENCE:
final String value = entries[i];
String columnName = namespace;
if (tag == GeneAssociationTag.EVIDENCE) columnName = columnName + EVIDENCE_SUFFIX;
else columnName = columnName + REFERENCE_SUFFIX;
List<String> valueList = row.getList(columnName, String.class);
if (valueList == null) valueList = new ArrayList<String>();
if (valueList.contains(value) == false) valueList.add(value);
row.set(columnName, valueList);
break;
case TAXON:
final String taxID = entries[i].split(":")[1];
final String taxName = speciesMap.get(taxID);
if (taxName != null) row.set(tag.toString(), taxName);
else if (taxID != null) row.set(tag.toString(), taxID);
break;
case ASPECT:
// Ignore these lines
break;
case DB_OBJECT_ID:
case DB_OBJECT_SYMBOL:
case DB_OBJECT_SYNONYM:
// Create consolidated id list attribute.
List<String> synList = row.getList(SYNONYM_COL_NAME, String.class);
if (synList == null) synList = new ArrayList<String>();
if (tag == GeneAssociationTag.DB_OBJECT_SYNONYM) {
final String[] vals = entries[i].split(LIST_DELIMITER);
for (String val : vals) {
if (synList.contains(val) == false) synList.add(val);
}
} else {
if (synList.contains(entries[i]) == false) synList.add(entries[i]);
}
row.set(SYNONYM_COL_NAME, synList);
break;
default:
if (LIST_INDEX.contains(i + 1)) {
final String[] vals = entries[i].split(LIST_DELIMITER);
List<String> listVals = row.getList(tag.toString(), String.class);
if (listVals == null) listVals = new ArrayList<String>();
for (String val : vals) {
if (listVals.contains(val) == false) listVals.add(val);
}
row.set(tag.toString(), listVals);
} else row.set(tag.toString(), entries[i]);
break;
}
}
}