org.biojava.nbio.structure.align.ClusterAltAligs Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of biojava-structure Show documentation
The protein structure modules of BioJava.
There is a newer version: 7.1.3
/*
 *                    BioJava development code
 *
 * This code may be freely distributed and modified under the
 * terms of the GNU Lesser General Public Licence.  This should
 * be distributed with the code.  If you do not have a copy,
 * see:
 *
 *      http://www.gnu.org/copyleft/lesser.html
 *
 * Copyright for this code is held jointly by the individual
 * authors.  These should be listed in @author doc comments.
 *
 * For more information on the BioJava project and its aims,
 * or to join the biojava-l mailing list, visit the home page
 * at:
 *
 *      http://www.biojava.org/
 *
 */
package org.biojava.nbio.structure.align;

import org.biojava.nbio.structure.align.pairwise.AlternativeAlignment;

import java.util.ArrayList;
import java.util.Arrays;
import java.util.Iterator;
import java.util.List;

/**
 * A class that clusters alternative alignments according to their
 * similarity.
 *
 * @author Andreas Prlic
 * @since 1.5
 * @version %I% %G%
 */
public class ClusterAltAligs {

	public static final int DEFAULT_CLUSTER_CUTOFF = 95;


	public static void cluster(AlternativeAlignment[] aligs ){
		cluster(aligs, DEFAULT_CLUSTER_CUTOFF);
	}

	public static void cluster(AlternativeAlignment[] aligs, int cutoff){


		List alist = Arrays.asList(aligs);
		List testAligs = new ArrayList<>(alist);

		List> clusters = new ArrayList<>();
		List excludeList = new ArrayList<>();

		// check how similar the eqrs are...
		for ( int i=0 ; i< aligs.length;i++){
			AlternativeAlignment a = aligs[i];
			if ( excludeList.contains(a)){
				continue;
			}
			int[] idxA = a.getIdx1();

			Iterator iter = testAligs.iterator();
			List remainList = new ArrayList<>();
			List currentCluster = new ArrayList<>();

			currentCluster.add(i);
			excludeList.add(a);

			int j=-1;
			while (iter.hasNext()){
				j++;
				AlternativeAlignment b = (AlternativeAlignment) iter.next();
				if ( excludeList.contains(b))
					continue;

				int[] idxB = b.getIdx1();

				// compare the eqrs..
				int samepos = 0;

				for ( int x = 0 ; x < idxA.length ;x++){
					int p1 =idxA[x];
					for (int y =0; y< idxB.length ; y++){
						int p2 = idxB[y];
						if ( p1 == p2){
							samepos++;
						}
					}
				}
				float perpos = (samepos / (float)idxA.length) * 100;
				//System.out.println("aa " + i + " samepos:"+ samepos +
				//		" l1:"+ idxA.length + " l2:" + idxB.length + " perpos:" + perpos);

				if ( perpos > cutoff){
					currentCluster.add(j);
					excludeList.add(b);
				} else {
					remainList.add(b);
				}

			}
			clusters.add(currentCluster);
			if ( remainList.size() == 0) {
				break;
			}
		}

		// now print the clusters...

		Iterator> iter = clusters.iterator();
		int cpos = 0;
		while (iter.hasNext()){
			cpos++;
			//System.out.println("cluster "+cpos+":");
			List cluster = iter.next();
			Iterator iter2 = cluster.iterator();
			while (iter2.hasNext()){
				Integer i = (Integer) iter2.next();

				AlternativeAlignment alig = aligs[i];
				alig.setCluster(cpos);
				//System.out.println( " ("+ aligs[i.intValue()]+")");

			}

		}
	}
}