de.julielab.geneexpbase.genemodel.GeneOrthologs Maven / Gradle / Ivy

Go to download
package de.julielab.geneexpbase.genemodel;

import com.google.common.collect.Sets;
import de.julielab.java.utilities.FileUtilities;
import de.julielab.java.utilities.cache.CacheAccess;
import de.julielab.java.utilities.cache.NoOpCacheAccess;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import javax.annotation.Nullable;
import javax.inject.Inject;
import javax.inject.Named;
import java.io.BufferedReader;
import java.io.File;
import java.io.IOException;
import java.io.Serializable;
import java.util.*;
import java.util.stream.Collectors;

import static de.julielab.geneexpbase.configuration.Configuration.GENE_ORTHOLOGS_PATH;

/**
 * This class models the contents of the gene_orthologs file.
 *
 * @see ftp://ftp.ncbi.nih.gov/gene/DATA/
 */
public class GeneOrthologs {
    private final static Logger log = LoggerFactory.getLogger(GeneOrthologs.class);
    private static GeneOrthologs instance;
    private Map> recordsByOtherGeneId;
    private Map> recordsByGeneId;
    private final CacheAccess> cache;
    private File geneOrthologsFile;

    @Inject
    public GeneOrthologs(@Nullable @Named(GENE_ORTHOLOGS_PATH) File geneOrthologsFile) {
        if (geneOrthologsFile != null) {
            log.info("Reading gene orthologs from {}", geneOrthologsFile);
            this.geneOrthologsFile = geneOrthologsFile;
        } else {
            log.warn("The path to the gene orthologs file was not specified. No ortholog information can be used.");
        }
        cache = new NoOpCacheAccess<>("", "");//CacheService.getInstance().getCacheAccess("Orthologs.db", "orthologs, ", CacheAccess.JAVA, CacheAccess.JAVA, mapSettings);
    }

    private synchronized void readOrthologsFile() {
        if (recordsByGeneId == null) {
            log.info("Reading gene orthologs from {}", geneOrthologsFile);
            try (BufferedReader br = FileUtilities.getReaderFromFile(geneOrthologsFile)) {
                List recordList = br.lines().filter(line -> !line.startsWith("#")).map(OrthologyRecord::new).collect(Collectors.toList());
                recordsByGeneId = recordList.stream().collect(Collectors.groupingBy(OrthologyRecord::getGeneId, HashMap::new, Collectors.toSet()));
                recordsByOtherGeneId = recordList.stream().collect(Collectors.groupingBy(OrthologyRecord::getOtherGeneId, HashMap::new, Collectors.toSet()));
            } catch (IOException e) {
                e.printStackTrace();
            }
            log.info("Done.");
        }
    }

    public Set getOrthologs(String geneId) {
        OrthologKey key = new OrthologKey(geneId);
        Set orthologyRecords = cache.get(key);
        if (orthologyRecords == null) {
            // the method reads file only once
            readOrthologsFile();
            orthologyRecords = new HashSet<>(Sets.union(recordsByGeneId.getOrDefault(geneId, Collections.emptySet()), recordsByOtherGeneId.getOrDefault(geneId, Collections.emptySet())));
            cache.put(key, orthologyRecords);
        }
        return orthologyRecords;
    }

    public Set getOrthologsInSpecies(String geneId, String targetTaxId) {
        OrthologKey key = new OrthologKey(geneId, targetTaxId);
        Set orthologyRecords = cache.get(key);
        if (orthologyRecords == null) {
            // the method reads the file only once
            readOrthologsFile();
            orthologyRecords = getOrthologs(geneId).stream().filter(r -> r.getTaxId().equals(targetTaxId) || r.getOtherTaxId().equals(targetTaxId)).collect(Collectors.toSet());
            cache.put(key, orthologyRecords);
        }
        return orthologyRecords;
    }

    private static class OrthologKey implements Serializable {
        private final String geneId;
        private final String targetTaxId;


        public OrthologKey(String geneId) {
            this(geneId, null);
        }

        public OrthologKey(String geneId, String targetTaxId) {
            this.geneId = geneId;
            this.targetTaxId = targetTaxId;
        }

        @Override
        public boolean equals(Object o) {
            if (this == o) return true;
            if (o == null || getClass() != o.getClass()) return false;
            OrthologKey that = (OrthologKey) o;
            return Objects.equals(geneId, that.geneId) &&
                    Objects.equals(targetTaxId, that.targetTaxId);
        }

        @Override
        public int hashCode() {
            return Objects.hash(geneId, targetTaxId);
        }
    }
}