zzh
/
gatk-3.8
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Now seeing results from the training data. There are still some critical problems in the quality of the output, but we're at least getting training output. 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@891 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
hanna 2009-06-03 20:41:07 +00:00
parent 4e41646c88
commit 8672ae6019
1 changed files with 18 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
java/src/org/broadinstitute/sting/playground/gatk/walkers/CovariateCounterWalker.java
View File

@ -112,7 +112,7 @@ public class CovariateCounterWalker extends LocusWalker<Integer, Integer> {
for (int i =0; i < reads.size(); i++ ) { for (int i =0; i < reads.size(); i++ ) {
SAMRecord read = reads.get(i); SAMRecord read = reads.get(i);
SAMReadGroupRecord readGroup = read.getHeader().getReadGroup((String)read.getAttribute("RG")); SAMReadGroupRecord readGroup = read.getHeader().getReadGroup((String)read.getAttribute("RG"));
if ( readGroup.getAttribute("PL") == "ILLUMINA" && if ( "ILLUMINA".equalsIgnoreCase(readGroup.getAttribute("PL").toString()) &&
!read.getReadNegativeStrandFlag() && !read.getReadNegativeStrandFlag() &&
(READ_GROUP.equals("none") || read.getAttribute("RG") != null && read.getAttribute("RG").equals(READ_GROUP)) && (READ_GROUP.equals("none") || read.getAttribute("RG") != null && read.getAttribute("RG").equals(READ_GROUP)) &&
(read.getMappingQuality() >= MIN_MAPPING_QUALITY) && (read.getMappingQuality() >= MIN_MAPPING_QUALITY) &&
@ -171,30 +171,32 @@ public class CovariateCounterWalker extends LocusWalker<Integer, Integer> {
} }
void writeTrainingData() { void writeTrainingData() {
PrintStream dinuc_out = null;
for (SAMReadGroupRecord readGroup : this.getToolkit().getEngine().getSAMHeader().getReadGroups()) { try {
for ( int dinuc_index=0; dinuc_index<NDINUCS; dinuc_index++) { dinuc_out = new PrintStream( OUTPUT_FILEROOT+".covariate_counts.csv");
PrintStream dinuc_out = null; dinuc_out.println("rg,dn,logitQ,pos,indicator,count");
try { for (SAMReadGroupRecord readGroup : this.getToolkit().getEngine().getSAMHeader().getReadGroups()) {
dinuc_out = new PrintStream( OUTPUT_FILEROOT+".covariate_counts.RG_"+readGroup.getReadGroupId()+"."+dinucIndex2bases(dinuc_index)+".csv"); for ( int dinuc_index=0; dinuc_index<NDINUCS; dinuc_index++) {
dinuc_out.println("logitQ,pos,indicator,count");
for ( RecalData datum: flattenData ) { for ( RecalData datum: flattenData ) {
if (string2dinucIndex(datum.dinuc) == dinuc_index) { if (string2dinucIndex(datum.dinuc) == dinuc_index) {
if ((datum.N - datum.B) > 0) if ((datum.N - datum.B) > 0)
dinuc_out.format("%d,%d,%d,%d\n", datum.qual, datum.pos, 0, datum.N - datum.B); dinuc_out.format("%s,%s,%d,%d,%d,%d%n", readGroup.getReadGroupId(), dinucIndex2bases(dinuc_index), datum.qual, datum.pos, 0, datum.N - datum.B);
if (datum.B > 0) if (datum.B > 0)
dinuc_out.format("%d,%d,%d,%d\n", datum.qual, datum.pos, 1, datum.B); dinuc_out.format("%s,%s,%d,%d,%d,%d%n", readGroup.getReadGroupId(), dinucIndex2bases(dinuc_index), datum.qual, datum.pos, 1, datum.B);
} }
} }
} catch (FileNotFoundException e) {
System.err.println("FileNotFoundException: " + e.getMessage());
} finally {
if (dinuc_out != null)
dinuc_out.close();
} }
} }
} }
catch (FileNotFoundException e) {
System.err.println("FileNotFoundException: " + e.getMessage());
return;
}
finally {
if (dinuc_out != null)
dinuc_out.close();
}
} }
class MeanReportedQuality { class MeanReportedQuality {