// For example, here is a line from the 5kb chr1 MAPQGE30 raw observed contact matrix
// (GM12878_combined/5kb_resolution_intrachromosomal/chr1/MAPQGE30/chr1_5kb.RAWobserved):
// 40000000 40100000 59.0
private static void processRawContactInformation(
String fileToRead,
double minValue,
ArrayList<DesiredChrContact> contactsToCheck,
boolean intra)
throws IOException {
// Check if sorted version is available
// If not make sorted available.
if (!Gpio.exists(fileToRead + ".sorted")) {
if (intra) {
umcg.genetica.io.chrContacts.SortIntraChrContacts.readNonSortedWriteSorted(
fileToRead, fileToRead + ".sorted");
} else {
umcg.genetica.io.chrContacts.SortInterChrContacts.readNonSortedWriteSorted(
fileToRead, fileToRead + ".sorted");
}
}
int numberToBeMatched = 0;
LineIterator it = FileUtils.lineIterator(new File(fileToRead + ".sorted"), "UTF-8");
try {
while (it.hasNext()) {
String[] parts = StringUtils.split(it.nextLine(), '\t');
int posChr1 = org.apache.commons.lang.math.NumberUtils.createInteger(parts[0]);
int posChr2 = org.apache.commons.lang.math.NumberUtils.createInteger(parts[1]);
while (numberToBeMatched < contactsToCheck.size()) {
if (posChr1 < contactsToCheck.get(numberToBeMatched).getChrLocationSmaller()) {
break;
} else if (posChr1 == contactsToCheck.get(numberToBeMatched).getChrLocationSmaller()) {
if (posChr2 < contactsToCheck.get(numberToBeMatched).getChrLocationLarger()) {
break;
}
if (posChr2 == contactsToCheck.get(numberToBeMatched).getChrLocationLarger()) {
double contact = org.apache.commons.lang.math.NumberUtils.createDouble(parts[2]);
if (contact >= minValue) {
contactsToCheck.get(numberToBeMatched).setContact();
numberToBeMatched++;
} else {
numberToBeMatched++;
}
} else if (posChr2 > contactsToCheck.get(numberToBeMatched).getChrLocationLarger()) {
numberToBeMatched++;
}
} else if (posChr1 > contactsToCheck.get(numberToBeMatched).getChrLocationSmaller()) {
numberToBeMatched++;
}
}
}
} finally {
LineIterator.closeQuietly(it);
}
}
// For example, here is a line from the 5kb chr1 MAPQGE30 raw observed contact matrix
// (GM12878_combined/5kb_resolution_intrachromosomal/chr1/MAPQGE30/chr1_5kb.RAWobserved):
// 40000000 40100000 59.0
// To normalize this entry using the KR normalization vector, one would divide 59.0 by the 8001st
// line ((40000000/5000)+1=8001) and the 8021st line ((40100000/5000)+1=8021)
// of GM12878_combined/5kb_resolution_intrachromosomal/chr1/MAPQGE30/chr1_5kb.KRnorm. The 8001st
// line of the KR norm file is 1.2988778370674694;
// The 8021st line of the KR norm file is 1.6080499717941548. So the corresponding KR normalized
// entry for the entry above is 59.0/(1.2988778370674694*1.6080499717941548)
// or 28.24776973966101.
// If the KR normalization vector file is empty or all NaNs, then the KR algorithm didn’t converge
// on that particular matrix (likely due to sparsity of the matrix).
private static void processNormalizedInterContactInformation(
String fileToRead,
String baseName,
String normMethod,
String chrSmaller,
String chrLarger,
ArrayList<DesiredChrContact> contactsToCheck,
String resolution,
double minValue,
TextFile outWriter)
throws IOException {
// ReadIn normalization chr1
TextFile inputNormChr1 =
new TextFile(
baseName + "\\chr" + chrSmaller + "_" + resolution + "." + normMethod, TextFile.R);
ArrayList<String> normFactorSmallerChr = inputNormChr1.readAsArrayList();
inputNormChr1.close();
// System.out.println("Done reading norm factor 1");
// ReadIn normalization chr2
TextFile inputNormChr2 =
new TextFile(
baseName + "\\chr" + chrLarger + "_" + resolution + "." + normMethod, TextFile.R);
ArrayList<String> normFactorLargerChr = inputNormChr2.readAsArrayList();
inputNormChr2.close();
// System.out.println("Done reading norm factor 2");
if (!Gpio.exists(fileToRead + ".sorted")) {
umcg.genetica.io.chrContacts.SortInterChrContacts.readNonSortedWriteSorted(
fileToRead, fileToRead + ".sorted");
}
int numberToBeMatched = 0;
LineIterator it = FileUtils.lineIterator(new File(fileToRead + ".sorted"), "UTF-8");
try {
while (it.hasNext()) {
String[] parts = StringUtils.split(it.nextLine(), '\t');
int posChr1 = org.apache.commons.lang.math.NumberUtils.createInteger(parts[0]);
int posChr2 = org.apache.commons.lang.math.NumberUtils.createInteger(parts[1]);
while (numberToBeMatched < contactsToCheck.size()) {
if (posChr1 < contactsToCheck.get(numberToBeMatched).getChrLocationSmaller()) {
break;
} else if (posChr1 == contactsToCheck.get(numberToBeMatched).getChrLocationSmaller()) {
if (posChr2 < contactsToCheck.get(numberToBeMatched).getChrLocationLarger()) {
break;
}
if (posChr2 == contactsToCheck.get(numberToBeMatched).getChrLocationLarger()) {
if (((posChr1 / getNumericResolution(resolution)) + 1)
> normFactorSmallerChr.size()) {
System.out.println(baseName);
System.out.println("Smaller");
System.out.println((posChr1 / getNumericResolution(resolution) + 1));
System.out.println(normFactorSmallerChr.size());
System.exit(-1);
}
if (((posChr2 / getNumericResolution(resolution)) + 1) > normFactorLargerChr.size()) {
System.out.println(baseName);
System.out.println("Larger");
System.out.println((posChr2 / getNumericResolution(resolution)) + 1);
System.out.println(normFactorLargerChr.size());
System.exit(-1);
}
String factor1Base =
normFactorSmallerChr.get((posChr1 / getNumericResolution(resolution)) + 1);
String factor2Base =
normFactorLargerChr.get((posChr2 / getNumericResolution(resolution)) + 1);
double factor1 = 1.0;
double factor2 = 1.0;
if (NumberUtils.isNumber(factor1Base) && NumberUtils.isNumber(factor2Base)) {
factor1 = Double.parseDouble(factor1Base);
factor2 = Double.parseDouble(factor2Base);
} else if (NumberUtils.isNumber(factor1Base)) {
factor1 = Double.parseDouble(factor1Base);
System.out.println("Error in files.");
System.out.println("Base 2 is reset to 1");
} else if (NumberUtils.isNumber(factor2Base)) {
factor2 = Double.parseDouble(factor2Base);
System.out.println("Error in files.");
System.out.println("Base 1 is reset to 1");
}
double contact =
org.apache.commons.lang.math.NumberUtils.createDouble(parts[2])
/ (factor1 * factor2);
if (contact >= minValue) {
outWriter.writeln(
contactsToCheck.get(numberToBeMatched).getSnpName()
+ "\t"
+ contactsToCheck.get(numberToBeMatched).getProbeName()
+ "\t"
+ posChr1
+ "\t"
+ posChr2
+ "\tContact\t"
+ contact
+ "\t"
+ org.apache.commons.lang.math.NumberUtils.createDouble(parts[2]));
numberToBeMatched++;
} else {
outWriter.writeln(
contactsToCheck.get(numberToBeMatched).getSnpName()
+ "\t"
+ contactsToCheck.get(numberToBeMatched).getProbeName()
+ "\t"
+ posChr1
+ "\t"
+ posChr2
+ "\t-\t-\t-");
numberToBeMatched++;
}
} else if (posChr2 > contactsToCheck.get(numberToBeMatched).getChrLocationLarger()) {
outWriter.writeln(
contactsToCheck.get(numberToBeMatched).getSnpName()
+ "\t"
+ contactsToCheck.get(numberToBeMatched).getProbeName()
+ "\t"
+ posChr1
+ "\t"
+ posChr2
+ "\t-\t-\t-");
numberToBeMatched++;
}
} else if (posChr1 > contactsToCheck.get(numberToBeMatched).getChrLocationSmaller()) {
outWriter.writeln(
contactsToCheck.get(numberToBeMatched).getSnpName()
+ "\t"
+ contactsToCheck.get(numberToBeMatched).getProbeName()
+ "\t"
+ posChr1
+ "\t"
+ posChr2
+ "\t-\t-\t-");
numberToBeMatched++;
}
}
}
} finally {
LineIterator.closeQuietly(it);
}
}