public final class AncestralTaxonomyInference {
private static final int MAX_CACHE_SIZE = 100000;
- private static final int MAX_TAXONOMIES_TO_RETURN = 100;
+ private static final int MAX_TAXONOMIES_TO_RETURN = 10;
private static final HashMap<String, UniProtTaxonomy> _sn_up_cache_map = new HashMap<String, UniProtTaxonomy>();
+ private static final HashMap<String, UniProtTaxonomy> _lineage_up_cache_map = new HashMap<String, UniProtTaxonomy>();
private static final HashMap<String, UniProtTaxonomy> _code_up_cache_map = new HashMap<String, UniProtTaxonomy>();
private static final HashMap<String, UniProtTaxonomy> _cn_up_cache_map = new HashMap<String, UniProtTaxonomy>();
private static final HashMap<String, UniProtTaxonomy> _id_up_cache_map = new HashMap<String, UniProtTaxonomy>();
if ( getSnTaxCacheMap().size() > MAX_CACHE_SIZE ) {
getSnTaxCacheMap().clear();
}
+ if ( getLineageTaxCacheMap().size() > MAX_CACHE_SIZE ) {
+ getLineageTaxCacheMap().clear();
+ }
if ( getCnTaxCacheMap().size() > MAX_CACHE_SIZE ) {
getCnTaxCacheMap().clear();
}
return _sn_up_cache_map;
}
+ synchronized private static HashMap<String, UniProtTaxonomy> getLineageTaxCacheMap() {
+ return _lineage_up_cache_map;
+ }
+
synchronized private static UniProtTaxonomy getTaxonomies( final HashMap<String, UniProtTaxonomy> cache,
- final String query,
- final QUERY_TYPE qt ) throws IOException {
+ final Object query,
+ final QUERY_TYPE qt ) throws IOException,
+ AncestralTaxonomyInferenceException {
if ( cache.containsKey( query ) ) {
return cache.get( query ).copy();
}
List<UniProtTaxonomy> up_taxonomies = null;
switch ( qt ) {
case ID:
- up_taxonomies = getTaxonomiesFromId( query );
+ up_taxonomies = getTaxonomiesFromId( ( String ) query );
break;
case CODE:
- up_taxonomies = getTaxonomiesFromTaxonomyCode( query );
+ up_taxonomies = getTaxonomiesFromTaxonomyCode( ( String ) query );
break;
case SN:
- up_taxonomies = getTaxonomiesFromScientificName( query );
+ up_taxonomies = getTaxonomiesFromScientificName( ( String ) query );
break;
case CN:
- up_taxonomies = getTaxonomiesFromCommonName( query );
+ up_taxonomies = getTaxonomiesFromCommonName( ( String ) query );
break;
+ case LIN:
+ return obtainUniProtTaxonomyFromLineage( ( List<String> ) query );
default:
throw new RuntimeException();
}
return UniProtWsTools.getTaxonomiesFromTaxonomyCode( query, MAX_TAXONOMIES_TO_RETURN );
}
- synchronized public static SortedSet<String> inferTaxonomyFromDescendents( final Phylogeny phy ) throws IOException {
+ synchronized public static void inferTaxonomyFromDescendents( final Phylogeny phy ) throws IOException,
+ AncestralTaxonomyInferenceException {
clearCachesIfTooLarge();
- final SortedSet<String> not_found = new TreeSet<String>();
for( final PhylogenyNodeIterator iter = phy.iteratorPostorder(); iter.hasNext(); ) {
final PhylogenyNode node = iter.next();
if ( !node.isExternal() ) {
- inferTaxonomyFromDescendents( node, not_found );
+ inferTaxonomyFromDescendents( node );
}
}
- return not_found;
}
- synchronized private static void inferTaxonomyFromDescendents( final PhylogenyNode n,
- final SortedSet<String> not_found )
- throws IOException {
+ synchronized private static void inferTaxonomyFromDescendents( final PhylogenyNode n ) throws IOException,
+ AncestralTaxonomyInferenceException {
if ( n.isExternal() ) {
throw new IllegalArgumentException( "attempt to infer taxonomy from descendants of external node" );
}
if ( desc.getNodeData().isHasTaxonomy()
&& ( isHasAppropriateId( desc.getNodeData().getTaxonomy() )
|| !ForesterUtil.isEmpty( desc.getNodeData().getTaxonomy().getScientificName() )
+ || !ForesterUtil.isEmpty( desc.getNodeData().getTaxonomy().getLineage() )
|| !ForesterUtil.isEmpty( desc.getNodeData().getTaxonomy().getTaxonomyCode() ) || !ForesterUtil
.isEmpty( desc.getNodeData().getTaxonomy().getCommonName() ) ) ) {
- final QUERY_TYPE qt = null;
- final String query = null;
- final UniProtTaxonomy up_tax = obtainUniProtTaxonomy( desc.getNodeData().getTaxonomy(), query, qt );
- String[] lineage = null;
- if ( up_tax != null ) {
- //lineage = obtainLineagePlusOwnScientificName( up_tax );
- lineage = up_tax.getLineageAsArray();
+ final UniProtTaxonomy up_tax = obtainUniProtTaxonomy( desc.getNodeData().getTaxonomy(), null, null );
+ if ( ( up_tax == null ) && ForesterUtil.isEmpty( desc.getNodeData().getTaxonomy().getLineage() ) ) {
+ String desc_str = "";
+ if ( !ForesterUtil.isEmpty( desc.getName() ) ) {
+ desc_str = "\"" + desc.getName() + "\"";
+ }
+ else {
+ desc_str = "[" + desc.getId() + "]";
+ }
+ System.out.println( desc.getNodeData().getTaxonomy().toString() );
+ System.out.println( ForesterUtil.stringListToString( desc.getNodeData().getTaxonomy().getLineage(),
+ " > " ) );
+ throw new AncestralTaxonomyInferenceException( "a taxonomy for node " + desc_str
+ + " could not be established from the database" );
}
+ String[] lineage = ForesterUtil.stringListToArray( desc.getNodeData().getTaxonomy().getLineage() );
if ( ( lineage == null ) || ( lineage.length < 1 ) ) {
- //TODO remove me
- System.out.println( "node " + desc.getNodeData().getTaxonomy().toString() + " has no lineage!" );
- not_found.add( desc.getNodeData().getTaxonomy().asText().toString() );
- return;
+ lineage = ForesterUtil.stringListToArray( up_tax.getLineage() );
+ }
+ if ( ( lineage == null ) || ( lineage.length < 1 ) ) {
+ throw new AncestralTaxonomyInferenceException( "a taxonomic lineage for node \""
+ + desc.getNodeData().getTaxonomy().toString() + "\" could not be established" );
}
if ( lineage.length < shortest_lin_length ) {
shortest_lin_length = lineage.length;
lineages.add( lineage );
}
else {
- String msg = "Node(s) with no or inappropriate taxonomic information found";
String node = "";
if ( !ForesterUtil.isEmpty( desc.getName() ) ) {
node = "\"" + desc.getName() + "\"";
else {
node = "[" + desc.getId() + "]";
}
- msg = "Node " + node + " has no or inappropriate taxonomic information";
- List<PhylogenyNode> e = desc.getAllExternalDescendants();
+ // final List<PhylogenyNode> e = desc.getAllExternalDescendants();
//TODO remove me!
- System.out.println();
- int x = 0;
- for( PhylogenyNode object : e ) {
- System.out.println( x + ":" );
- System.out.println( object.getName() + " " );
- x++;
- }
- System.out.println();
+ // System.out.println();
+ // int x = 0;
+ // for( final PhylogenyNode object : e ) {
+ // System.out.println( x + ":" );
+ // System.out.println( object.getName() + " " );
+ // x++;
+ // }
+ // System.out.println();
//
- throw new IllegalArgumentException( msg );
+ throw new AncestralTaxonomyInferenceException( "node " + node
+ + " has no or inappropriate taxonomic information" );
}
}
- String last_common_lineage = null;
+ final List<String> last_common_lineage = new ArrayList<String>();
+ String last_common = null;
if ( shortest_lin_length > 0 ) {
I: for( int i = 0; i < shortest_lin_length; ++i ) {
final String lineage_0 = lineages.get( 0 )[ i ];
break I;
}
}
- last_common_lineage = lineage_0;
+ last_common_lineage.add( lineage_0 );
+ last_common = lineage_0;
}
}
- if ( last_common_lineage == null ) {
- System.out.println( "No common lineage for:" );
- int counter = 0;
- for( String[] strings : lineages ) {
- System.out.print( counter + ": " );
- ++counter;
- for( String string : strings ) {
- System.out.print( string + " " );
+ if ( last_common_lineage.isEmpty() ) {
+ boolean saw_viruses = false;
+ boolean saw_cellular_organism = false;
+ for( final String[] lineage : lineages ) {
+ if ( lineage.length > 0 ) {
+ if ( lineage[ 0 ].equalsIgnoreCase( UniProtTaxonomy.VIRUSES ) ) {
+ saw_viruses = true;
+ }
+ else if ( lineage[ 0 ].equalsIgnoreCase( UniProtTaxonomy.CELLULAR_ORGANISMS ) ) {
+ saw_cellular_organism = true;
+ }
+ if ( saw_cellular_organism && saw_viruses ) {
+ break;
+ }
+ }
+ }
+ if ( saw_cellular_organism && saw_viruses ) {
+ last_common_lineage.add( UniProtTaxonomy.CELLULAR_ORGANISMS );
+ last_common = UniProtTaxonomy.CELLULAR_ORGANISMS;
+ }
+ else {
+ String msg = "no common lineage for:\n";
+ int counter = 0;
+ for( final String[] strings : lineages ) {
+ msg += counter + ": ";
+ ++counter;
+ for( final String string : strings ) {
+ msg += string + " ";
+ }
+ msg += "\n";
}
- System.out.println();
+ throw new AncestralTaxonomyInferenceException( msg );
}
- return;
}
- // if ( !n.getNodeData().isHasTaxonomy() ) {
- // n.getNodeData().setTaxonomy( new Taxonomy() );
- // }
final Taxonomy tax = new Taxonomy();
n.getNodeData().setTaxonomy( tax );
- tax.setScientificName( last_common_lineage );
- final UniProtTaxonomy up_tax = obtainUniProtTaxonomyFromSn( last_common_lineage );
+ tax.setScientificName( last_common );
+ final UniProtTaxonomy up_tax = obtainUniProtTaxonomyFromLineage( last_common_lineage );
if ( up_tax != null ) {
if ( !ForesterUtil.isEmpty( up_tax.getRank() ) ) {
try {
if ( !ForesterUtil.isEmpty( up_tax.getSynonym() ) && !tax.getSynonyms().contains( up_tax.getSynonym() ) ) {
tax.getSynonyms().add( up_tax.getSynonym() );
}
+ if ( up_tax.getLineage() != null ) {
+ tax.setLineage( new ArrayList<String>() );
+ for( final String lin : up_tax.getLineage() ) {
+ if ( !ForesterUtil.isEmpty( lin ) ) {
+ tax.getLineage().add( lin );
+ }
+ }
+ }
+ }
+ if ( ForesterUtil.isEmpty( tax.getLineage() ) ) {
+ tax.setLineage( new ArrayList<String>() );
+ for( final String lin : last_common_lineage ) {
+ if ( !ForesterUtil.isEmpty( lin ) ) {
+ tax.getLineage().add( lin );
+ }
+ }
}
for( final PhylogenyNode desc : descs ) {
if ( !desc.isExternal() && desc.getNodeData().isHasTaxonomy()
synchronized public static SortedSet<String> obtainDetailedTaxonomicInformation( final Phylogeny phy,
final boolean delete )
- throws IOException {
+ throws IOException, AncestralTaxonomyInferenceException {
clearCachesIfTooLarge();
final SortedSet<String> not_found = new TreeSet<String>();
List<PhylogenyNode> not_found_external_nodes = null;
not_found_external_nodes.add( node );
}
}
- UniProtTaxonomy up_tax = null;
+ UniProtTaxonomy uniprot_tax = null;
if ( ( tax != null )
&& ( isHasAppropriateId( tax ) || !ForesterUtil.isEmpty( tax.getScientificName() )
|| !ForesterUtil.isEmpty( tax.getTaxonomyCode() ) || !ForesterUtil.isEmpty( tax
.getCommonName() ) ) ) {
- up_tax = obtainUniProtTaxonomy( tax, null, qt );
- if ( up_tax != null ) {
- updateTaxonomy( qt, node, tax, up_tax );
+ uniprot_tax = obtainUniProtTaxonomy( tax, null, qt );
+ if ( uniprot_tax != null ) {
+ updateTaxonomy( qt, node, tax, uniprot_tax );
}
else {
not_found.add( tax.toString() );
}
}
if ( delete ) {
- for( PhylogenyNode node : not_found_external_nodes ) {
- phy.deleteSubtree( node, false );
+ for( final PhylogenyNode node : not_found_external_nodes ) {
+ phy.deleteSubtree( node, true );
}
+ phy.externalNodesHaveChanged();
+ phy.hashIDs();
phy.recalculateNumberOfExternalDescendants( true );
}
return not_found;
}
- // TODO this might not be needed anymore
- // synchronized private static String[] obtainLineagePlusOwnScientificName( final UniProtTaxonomy up_tax ) {
- // final String[] lineage = up_tax.getLineageAsArray();
- // final String[] lin_plus_self = new String[ lineage.length + 1 ];
- // for( int i = 0; i < lineage.length; ++i ) {
- // lin_plus_self[ i ] = lineage[ i ];
- // }
- // lin_plus_self[ lineage.length ] = up_tax.getScientificName();
- // return lin_plus_self;
- // }
- synchronized private static UniProtTaxonomy obtainUniProtTaxonomy( final Taxonomy tax, String query, QUERY_TYPE qt )
- throws IOException {
+ synchronized public static UniProtTaxonomy obtainUniProtTaxonomy( final Taxonomy tax, Object query, QUERY_TYPE qt )
+ throws IOException, AncestralTaxonomyInferenceException {
if ( isHasAppropriateId( tax ) ) {
query = tax.getIdentifier().getValue();
qt = QUERY_TYPE.ID;
return getTaxonomies( getIdTaxCacheMap(), query, qt );
}
else if ( !ForesterUtil.isEmpty( tax.getScientificName() ) ) {
- query = tax.getScientificName();
- qt = QUERY_TYPE.SN;
- return getTaxonomies( getSnTaxCacheMap(), query, qt );
+ if ( !ForesterUtil.isEmpty( tax.getLineage() ) ) {
+ query = tax.getLineage();
+ qt = QUERY_TYPE.LIN;
+ return getTaxonomies( getLineageTaxCacheMap(), query, qt );
+ }
+ else {
+ query = tax.getScientificName();
+ qt = QUERY_TYPE.SN;
+ return getTaxonomies( getSnTaxCacheMap(), query, qt );
+ }
}
else if ( !ForesterUtil.isEmpty( tax.getTaxonomyCode() ) ) {
query = tax.getTaxonomyCode();
}
}
- synchronized private static UniProtTaxonomy obtainUniProtTaxonomyFromSn( final String sn ) throws IOException {
+ synchronized private static UniProtTaxonomy obtainUniProtTaxonomyFromLineage( final List<String> lineage )
+ throws AncestralTaxonomyInferenceException, IOException {
+ final String lineage_str = ForesterUtil.stringListToString( lineage, ">" );
UniProtTaxonomy up_tax = null;
- if ( getSnTaxCacheMap().containsKey( sn ) ) {
- up_tax = getSnTaxCacheMap().get( sn ).copy();
+ if ( getLineageTaxCacheMap().containsKey( lineage_str ) ) {
+ up_tax = getLineageTaxCacheMap().get( lineage_str ).copy();
}
else {
- final List<UniProtTaxonomy> up_taxonomies = getTaxonomiesFromScientificName( sn );
- if ( ( up_taxonomies != null ) && ( up_taxonomies.size() == 1 ) ) {
- up_tax = up_taxonomies.get( 0 );
- getSnTaxCacheMap().put( sn, up_tax );
+ final List<UniProtTaxonomy> up_taxonomies = getTaxonomiesFromScientificName( lineage
+ .get( lineage.size() - 1 ) );
+ if ( ( up_taxonomies != null ) && ( up_taxonomies.size() > 0 ) ) {
+ for( final UniProtTaxonomy up_taxonomy : up_taxonomies ) {
+ boolean match = true;
+ I: for( int i = 0; i < lineage.size(); ++i ) {
+ if ( !lineage.get( i ).equalsIgnoreCase( up_taxonomy.getLineage().get( i ) ) ) {
+ match = false;
+ break I;
+ }
+ }
+ if ( match ) {
+ if ( up_tax != null ) {
+ throw new AncestralTaxonomyInferenceException( "lineage \""
+ + ForesterUtil.stringListToString( lineage, " > " ) + "\" is not unique" );
+ }
+ up_tax = up_taxonomy;
+ }
+ }
+ if ( up_tax == null ) {
+ throw new AncestralTaxonomyInferenceException( "lineage \""
+ + ForesterUtil.stringListToString( lineage, " > " ) + "\" not found" );
+ }
+ getLineageTaxCacheMap().put( lineage_str, up_tax );
+ if ( !ForesterUtil.isEmpty( up_tax.getScientificName() ) ) {
+ getSnTaxCacheMap().put( up_tax.getScientificName(), up_tax );
+ }
if ( !ForesterUtil.isEmpty( up_tax.getCode() ) ) {
getCodeTaxCacheMap().put( up_tax.getCode(), up_tax );
}
&& ForesterUtil.isEmpty( tax.getScientificName() ) ) {
tax.setScientificName( up_tax.getScientificName() );
}
- if ( node.isExternal()
- && ( ( qt != QUERY_TYPE.CODE ) && !ForesterUtil.isEmpty( up_tax.getCode() ) && ForesterUtil
- .isEmpty( tax.getTaxonomyCode() ) ) ) {
+ if ( node.isExternal() && ( qt != QUERY_TYPE.CODE ) && !ForesterUtil.isEmpty( up_tax.getCode() )
+ && ForesterUtil.isEmpty( tax.getTaxonomyCode() ) ) {
tax.setTaxonomyCode( up_tax.getCode() );
}
if ( ( qt != QUERY_TYPE.CN ) && !ForesterUtil.isEmpty( up_tax.getCommonName() )
tax.setRank( "" );
}
}
- if ( ( qt != QUERY_TYPE.ID ) && !ForesterUtil.isEmpty( up_tax.getId() ) && ( tax.getIdentifier() == null ) ) {
+ if ( ( qt != QUERY_TYPE.ID ) && !ForesterUtil.isEmpty( up_tax.getId() )
+ && ( ( tax.getIdentifier() == null ) || ForesterUtil.isEmpty( tax.getIdentifier().getValue() ) ) ) {
tax.setIdentifier( new Identifier( up_tax.getId(), "uniprot" ) );
}
+ if ( up_tax.getLineage() != null ) {
+ tax.setLineage( new ArrayList<String>() );
+ for( final String lin : up_tax.getLineage() ) {
+ if ( !ForesterUtil.isEmpty( lin ) ) {
+ tax.getLineage().add( lin );
+ }
+ }
+ }
}
private enum QUERY_TYPE {
- CODE, SN, CN, ID;
+ CODE, SN, CN, ID, LIN;
}
}