Package htsjdk.samtools.util

Examples of htsjdk.samtools.util.FormatUtil

53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
        { 15.0d, 23, 55, 67.88888}
    };

    @Test(dataProvider = "basicInputParserData")
    public void testTextFileParser(Object fileOrStream) throws IOException {
        FormatUtil format = new FormatUtil();

        List<String> expected = new ArrayList<String>();
        if (fileOrStream instanceof File) {
            BufferedReader reader = new BufferedReader(new FileReader((File)fileOrStream));
            String line = null;
            while ((line = reader.readLine()) != null)  {
                if (!line.startsWith("#")) expected.add(line);
            }
            reader.close();
        }

        BasicInputParser parser = fileOrStream instanceof File
            ? new BasicInputParser(true, (File)fileOrStream )
            : new BasicInputParser(true, (InputStream)fileOrStream);
        int index = 0;
        while (parser.hasNext())
        {
            String parts[] = parser.next();
            if (fileOrStream instanceof File) {
                // Can't have the parser and the reader workking with an InputStream at the same time
                // so we only do this test with the file
                Assert.assertEquals(parser.getCurrentLine(), expected.get(index));
            }
            // Line 4 is a comment, so there's a gap in the line numbers
            Assert.assertEquals(parser.getCurrentLineNumber(), index <= 2 ? index+1 : index+2);
            Assert.assertEquals(parts.length, 4);
            if (index < 4) {
                for (int i = 0; i < parts.length; i++) {
                    Assert.assertEquals(parts, testFile1Data[index]);
                }
            }
            else {
                Assert.assertEquals(testFile1Data[index][0], format.parseDouble(parts[0]));
                Assert.assertEquals(testFile1Data[index][1], format.parseInt(parts[1]));
                Assert.assertEquals(testFile1Data[index][2], format.parseInt(parts[2]));
                Assert.assertEquals(testFile1Data[index][3], format.parseDouble(parts[3]));
            }
            index++;
        }
    }
View Full Code Here
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
        };
    }

    @Test(dataProvider = "multiFileParsingData")
    public void testMultiFileParsing(Object fileOrStream1, Object fileOrStream2) throws IOException {
        FormatUtil format = new FormatUtil();

        List<String> expected = new ArrayList<String>();
        if (fileOrStream1 instanceof File) {
            BufferedReader reader = new BufferedReader(new FileReader((File)fileOrStream1));
            String line = null;
            while ((line = reader.readLine()) != null)  {
                if (!line.startsWith("#")) expected.add(line);
            }
            reader.close();
            reader = new BufferedReader(new FileReader((File)fileOrStream2));
            while ((line = reader.readLine()) != null)  {
                if (!line.startsWith("#")) expected.add(line);
            }
            reader.close();
        }

        BasicInputParser parser = fileOrStream1 instanceof File
            ? new BasicInputParser(true, (File)fileOrStream1, (File)fileOrStream2 )
            : new BasicInputParser(true, (InputStream)fileOrStream1, (InputStream)fileOrStream2);
        int index = 0;
        // Line 4 is a comment, so there's a gap in the line numbers
        int expectedLineNumbers[] = {1,2,3,5,6,1,2,3,5,6};
        while (parser.hasNext())
        {
            String parts[] = parser.next();
            if (fileOrStream1 instanceof File) {
                // Can't have the parser and the reader working with an InputStream at the same time
                // so we only test the files
                Assert.assertEquals(parser.getCurrentLine(), expected.get(index));
            }
            Assert.assertEquals(parser.getCurrentLineNumber(), expectedLineNumbers[index]);
            Assert.assertEquals(parts.length, 4);
            int indexIntoTestData = (index<5) ? index : index-5;
            if (index != 4 && index != 9) {
                for (int i = 0; i < parts.length; i++) {
                    Assert.assertEquals(parts, testFile1Data[indexIntoTestData]);
                }
            }
            else {
                Assert.assertEquals(testFile1Data[indexIntoTestData][0], format.parseDouble(parts[0]));
                Assert.assertEquals(testFile1Data[indexIntoTestData][1], format.parseInt(parts[1]));
                Assert.assertEquals(testFile1Data[indexIntoTestData][2], format.parseInt(parts[2]));
                Assert.assertEquals(testFile1Data[indexIntoTestData][3], format.parseDouble(parts[3]));
            }
            index++;
        }
    }
View Full Code Here
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
        private void calculateGcMetrics() {
            if (this.intervalToGc != null) {
                log.info("Calculating GC metrics");

                // Setup the output file if we're outputting per-target coverage
                FormatUtil fmt = new FormatUtil();
                final PrintWriter out;
                try {
                    if (perTargetOutput != null) {
                        out = new PrintWriter(perTargetOutput);
                        out.println("chrom\tstart\tend\tlength\tname\t%gc\tmean_coverage\tnormalized_coverage");
                    }
                    else {
                        out = null;
                    }
                }
                catch (IOException ioe) { throw new RuntimeIOException(ioe); }

                final int bins = 101;
                final long[] targetBasesByGc  = new long[bins];
                final long[] alignedBasesByGc = new long[bins];

                for (final Map.Entry<Interval,Coverage> entry : this.coverageByTarget.entrySet()) {
                    final Interval interval = entry.getKey();
                    final Coverage cov = entry.getValue();

                    final double gcDouble = this.intervalToGc.get(interval);
                    final int gc = (int) Math.round(gcDouble * 100);

                    targetBasesByGc[gc]  += interval.length();
                    alignedBasesByGc[gc] += cov.getTotal();

                    if (out != null) {
                        final double coverage = cov.getTotal() / (double) interval.length();

                        out.println(interval.getSequence() + "\t" +
                                    interval.getStart() + "\t" +
                                    interval.getEnd() + "\t" +
                                    interval.length() + "\t" +
                                    interval.getName() + "\t" +
                                    fmt.format(gcDouble) + "\t" +
                                    fmt.format(coverage) + "\t" +
                                    fmt.format(coverage / this.metrics.MEAN_TARGET_COVERAGE)
                        );
                    }
                }

                if (out != null) out.close();
View Full Code Here
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
    /**
     * Fill in expected duplication metrics directly from the input records given to this tester
     */
    private void updateExpectedDuplicationMetrics() {

        final FormatUtil formatter = new FormatUtil();

        final CloseableIterator<SAMRecord> inputRecordIterator = this.getRecordIterator();
        while (inputRecordIterator.hasNext()) {
            final SAMRecord record = inputRecordIterator.next();
            if (!record.isSecondaryOrSupplementary()) {
                final String key = samRecordToDuplicatesFlagsKey(record);
                if (!this.duplicateFlags.containsKey(key)) {
                    System.err.println("DOES NOT CONTAIN KEY: " + key);
                }
                final boolean isDuplicate = this.duplicateFlags.get(key);

                // First bring the simple metricsFile up to date
                if (record.getReadUnmappedFlag()) {
                    ++expectedMetrics.UNMAPPED_READS;
                }
                else if (!record.getReadPairedFlag() || record.getMateUnmappedFlag()) {
                    ++expectedMetrics.UNPAIRED_READS_EXAMINED;
                    if (isDuplicate) ++expectedMetrics.UNPAIRED_READ_DUPLICATES;
                }
                else {
                    ++expectedMetrics.READ_PAIRS_EXAMINED; // will need to be divided by 2 at the end
                    if (isDuplicate) ++expectedMetrics.READ_PAIR_DUPLICATES; // will need to be divided by 2 at the end
                }
            }
        }
        expectedMetrics.READ_PAIR_DUPLICATES = expectedMetrics.READ_PAIR_DUPLICATES / 2;
        expectedMetrics.READ_PAIRS_EXAMINED = expectedMetrics.READ_PAIRS_EXAMINED / 2;
        expectedMetrics.calculateDerivedMetrics();

        // Have to run this Double value through the same format/parsing operations as during a file write/read
        expectedMetrics.PERCENT_DUPLICATION = formatter.parseDouble(formatter.format(expectedMetrics.PERCENT_DUPLICATION));
    }
View Full Code Here
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
        nonZeroCounts.put(new TruthAndCallStates(TruthState.VC_FILTERED, CallState.VC_FILTERED), 40);

        GenotypeConcordanceCounts concordanceCounts = genotypeConcordance.getSnpCounter();
        assertNonZeroCountsAgree(concordanceCounts, nonZeroCounts);

        final FormatUtil fmt = new FormatUtil();

        final GenotypeConcordanceScheme scheme = new GenotypeConcordanceScheme();
        concordanceCounts.validateCountsAgainstScheme(scheme);

        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.HET_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.HET_CALL_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.HET_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.HOM_VAR_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.HOM_VAR_CALL_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.HOM_VAR_TRUTH_STATES)), "?");
        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.VAR_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.VAR_CALL_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.VAR_TRUTH_STATES)), "1");

        // Now run it again with different samples
        genotypeConcordance.TRUTH_VCF = CEU_TRIOS_SNPS_VCF;
        genotypeConcordance.TRUTH_SAMPLE = "NA12878";
        genotypeConcordance.CALL_VCF = CEU_TRIOS_SNPS_VCF;
        genotypeConcordance.CALL_SAMPLE = "NA12891";
        Assert.assertEquals(genotypeConcordance.instanceMain(new String[0]), 0);

        nonZeroCounts.clear();
        nonZeroCounts.put(new TruthAndCallStates(TruthState.HOM_REF, CallState.HET_REF_VAR1), 31);
        nonZeroCounts.put(new TruthAndCallStates(TruthState.HET_REF_VAR1, CallState.HOM_REF), 30);
        nonZeroCounts.put(new TruthAndCallStates(TruthState.HET_REF_VAR1, CallState.HET_REF_VAR1), 50);
        nonZeroCounts.put(new TruthAndCallStates(TruthState.HET_REF_VAR1, CallState.HOM_VAR1), 24);
        nonZeroCounts.put(new TruthAndCallStates(TruthState.HOM_VAR1, CallState.HET_REF_VAR1), 18);
        nonZeroCounts.put(new TruthAndCallStates(TruthState.HOM_VAR1, CallState.HOM_VAR1), 41);
        nonZeroCounts.put(new TruthAndCallStates(TruthState.VC_FILTERED, CallState.VC_FILTERED), 49);

        concordanceCounts = genotypeConcordance.getSnpCounter();
        assertNonZeroCountsAgree(concordanceCounts, nonZeroCounts);

        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.HET_TRUTH_STATES)), "0.711538");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.HET_CALL_STATES)), "0.686869");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.HET_TRUTH_STATES)), "0.769231");
        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.HOM_VAR_TRUTH_STATES)), "0.766234");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.HOM_VAR_CALL_STATES)), "0.730337");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.HOM_VAR_TRUTH_STATES)), "?");
        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.VAR_TRUTH_STATES)), "0.734807");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.VAR_CALL_STATES)), "0.707447");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.VAR_TRUTH_STATES)), "0.769231");
    }
View Full Code Here
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
        nonZeroCounts.put(new TruthAndCallStates(TruthState.VC_FILTERED, CallState.VC_FILTERED), 2);

        GenotypeConcordanceCounts concordanceCounts = genotypeConcordance.getSnpCounter();
        assertNonZeroCountsAgree(concordanceCounts, nonZeroCounts);

        final FormatUtil fmt = new FormatUtil();

        final GenotypeConcordanceScheme scheme = new GenotypeConcordanceScheme();
        concordanceCounts.validateCountsAgainstScheme(scheme);

        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.HET_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.HET_CALL_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.HET_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.HOM_VAR_TRUTH_STATES)), "?");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.HOM_VAR_CALL_STATES)), "?");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.HOM_VAR_TRUTH_STATES)), "?");
        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.VAR_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.VAR_CALL_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.VAR_TRUTH_STATES)), "1");

        // Now run it again with different samples
        genotypeConcordance.TRUTH_VCF = CEU_TRIOS_SNPS_VCF;
        genotypeConcordance.TRUTH_SAMPLE = "NA12878";
        genotypeConcordance.CALL_VCF = CEU_TRIOS_SNPS_VCF;
        genotypeConcordance.CALL_SAMPLE = "NA12891";
        genotypeConcordance.INTERVALS = Collections.singletonList(INTERVALS_FILE);
        Assert.assertEquals(genotypeConcordance.instanceMain(new String[0]), 0);

        nonZeroCounts.clear();
        nonZeroCounts.put(new TruthAndCallStates(TruthState.HOM_REF, CallState.HET_REF_VAR1), 1);
        nonZeroCounts.put(new TruthAndCallStates(TruthState.HET_REF_VAR1, CallState.HET_REF_VAR1), 1);
        nonZeroCounts.put(new TruthAndCallStates(TruthState.VC_FILTERED, CallState.VC_FILTERED), 2);

        concordanceCounts = genotypeConcordance.getSnpCounter();
        assertNonZeroCountsAgree(concordanceCounts, nonZeroCounts);

        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.HET_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.HET_CALL_STATES)), "0.5");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.HET_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.HOM_VAR_TRUTH_STATES)), "?");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.HOM_VAR_CALL_STATES)), "?");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.HOM_VAR_TRUTH_STATES)), "?");
        Assert.assertEquals(fmt.format(concordanceCounts.getSensitivity(scheme, GenotypeConcordanceCounts.VAR_TRUTH_STATES)), "1");
        Assert.assertEquals(fmt.format(concordanceCounts.Ppv(scheme, GenotypeConcordanceCounts.VAR_CALL_STATES)), "0.5");
        Assert.assertEquals(fmt.format(concordanceCounts.getSpecificity(scheme, GenotypeConcordanceCounts.VAR_TRUTH_STATES)), "1");
    }
View Full Code Here
TOP

Related Classes of htsjdk.samtools.util.FormatUtil

  • picard.analysis.directed.TargetMetricsCollector$PerUnitTargetMetricCollector
  • picard.sam.markduplicates.AbstractMarkDuplicatesCommandLineProgramTester
  • picard.util.TextFileParsersTest
  • picard.vcf.GenotypeConcordanceTest
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.