comparison src/edu/unc/genomics/nucleosomes/NRLCalculator.java @ 2:e16016635b2a

Uploaded
author timpalpant
date Mon, 13 Feb 2012 22:12:06 -0500
parents
children
comparison
equal deleted inserted replaced
1:a54db233ee3d 2:e16016635b2a
1 package edu.unc.genomics.nucleosomes;
2
3 import java.io.BufferedWriter;
4 import java.io.IOException;
5 import java.nio.charset.Charset;
6 import java.nio.file.Files;
7 import java.nio.file.Path;
8 import java.util.ArrayList;
9 import java.util.Collections;
10 import java.util.HashMap;
11 import java.util.List;
12 import java.util.Map;
13
14 import org.apache.log4j.Logger;
15
16 import com.beust.jcommander.Parameter;
17
18 import edu.unc.genomics.CommandLineTool;
19 import edu.unc.genomics.Interval;
20 import edu.unc.genomics.ReadablePathValidator;
21 import edu.unc.genomics.io.IntervalFile;
22
23 public class NRLCalculator extends CommandLineTool {
24
25 private static final Logger log = Logger.getLogger(NRLCalculator.class);
26
27 @Parameter(names = {"-i", "--input"}, description = "Input file (nucleosome calls)", required = true, validateWith = ReadablePathValidator.class)
28 public Path inputFile;
29 @Parameter(names = {"-l", "--loci"}, description = "Genomic loci (Bed format)", required = true)
30 public IntervalFile<? extends Interval> lociFile;
31 @Parameter(names = {"-o", "--output"}, description = "Output file (NRL for each gene)", required = true)
32 public Path outputFile;
33
34 private Map<String,List<NucleosomeCall>> nucs = new HashMap<>();
35
36 private List<NucleosomeCall> getIntervalNucleosomes(Interval i) {
37 List<NucleosomeCall> intervalNucs = new ArrayList<>();
38 for (NucleosomeCall call : nucs.get(i.getChr())) {
39 if (call.getDyad() >= i.low() && call.getDyad() <= i.high()) {
40 intervalNucs.add(call);
41 }
42 }
43
44 return intervalNucs;
45 }
46
47 @Override
48 public void run() throws IOException {
49 log.debug("Initializing input file");
50 NucleosomeCallsFile nucsFile = new NucleosomeCallsFile(inputFile);
51 log.debug("Loading all nucleosomes");
52 for (NucleosomeCall nuc : nucsFile) {
53 if (nuc == null) continue;
54 if (!nucs.containsKey(nuc.getChr())) {
55 nucs.put(nuc.getChr(), new ArrayList<NucleosomeCall>());
56 }
57 nucs.get(nuc.getChr()).add(nuc);
58 }
59 nucsFile.close();
60
61 log.debug("Initializing output file");
62 try (BufferedWriter writer = Files.newBufferedWriter(outputFile, Charset.defaultCharset())) {
63 log.debug("Calculating nucleosome spacing for each interval");
64 NucleosomeCall.DyadComparator comparator = new NucleosomeCall.DyadComparator();
65 for (Interval interval : lociFile) {
66 writer.write(interval.toBed());
67
68 // Get all of the nucleosomes within this interval
69 List<NucleosomeCall> intervalNucs = getIntervalNucleosomes(interval);
70
71 if (intervalNucs.size() > 1) {
72 // Sort the list by nucleosome position
73 Collections.sort(intervalNucs, comparator);
74 if (interval.isCrick()) {
75 Collections.reverse(intervalNucs);
76 }
77
78 for (int i = 1; i < Math.min(intervalNucs.size(), 10); i++) {
79 writer.write("\t" + Math.abs(intervalNucs.get(i).getDyad()-intervalNucs.get(i-1).getDyad()));
80 }
81 }
82
83 writer.newLine();
84 }
85 }
86
87 lociFile.close();
88 }
89
90 public static void main(String[] args) throws IOException {
91 new NRLCalculator().instanceMain(args);
92 }
93 }