/** @param args the command line arguments */
@SuppressWarnings("ManualArrayToCollectionCopy")
public static void main(String[] args) throws Exception {
// eQTLTextFile eQTLsTextFile = new
// QTLTextFile("D:\\UMCG\\Genetica\\Projects\\RnaSeqEqtl\\batch9_eQTLmapping\\result_non-geuvadis_maf0.05_call0.5_pcs100_normalizedPCA_meta\\notInGeuvadis.txt", false);
// eQTLTextFile eQTLsTextFile = new
// QTLTextFile("D:\\UMCG\\Genetica\\Projects\\RnaSeqEqtl\\batch9_eQTLmapping\\result_all_maf0.05_call0.5_pcs100_normalizedPCA_meta_specialPermutation_fix\\eQTLsFDR0.05-ProbeLevel.txt", false);
QTLTextFile eQTLsTextFile =
new QTLTextFile(
"D:\\UMCG\\Genetica\\Projects\\RnaSeqEqtl\\batch9_eQTLmapping\\result_geuvadis_maf0.05_call0.5_pcs100_normalizedPCA_meta_fix\\eQTLsFDR0.05-ProbeLevel.txt",
false);
BufferedReader aseReader =
new BufferedReader(
new FileReader(
"D:\\UMCG\\Genetica\\Projects\\RnaSeqEqtl\\Ase\\geuvadis_maskAll4_r20_a10_p2_s5_rq17_m1_gatkGenoGq30\\ase_bh.txt"));
HashMap<String, ArrayList<EQTL>> eQtls = new HashMap<String, ArrayList<EQTL>>();
for (Iterator<EQTL> eQtlIt = eQTLsTextFile.getEQtlIterator(); eQtlIt.hasNext(); ) {
EQTL eQtl = eQtlIt.next();
String eQtlKey = eQtl.getRsChr() + ":" + eQtl.getRsChrPos();
ArrayList<EQTL> posEqtls = eQtls.get(eQtlKey);
if (posEqtls == null) {
posEqtls = new ArrayList<EQTL>(1);
eQtls.put(eQtlKey, posEqtls);
}
posEqtls.add(eQtl);
}
int aseTotal = 0;
int aseWithEQtl = 0;
int sameDirection = 0;
int oppositeDirection = 0;
HashSet<String> countedGenes = new HashSet<String>();
aseReader.readLine(); // header
String line;
String[] elements;
while ((line = aseReader.readLine()) != null) {
elements = TAB_PATTERN.split(line);
HashSet<String> aseGenes = new HashSet<String>();
for (String gene : COMMA_PATTERN.split(elements[ASE_GENES_COLUMN])) {
aseGenes.add(gene);
}
++aseTotal;
ArrayList<EQTL> posEqtls =
eQtls.get(elements[ASE_CHR_COLUMN] + ":" + elements[ASE_POS_COLUMN]);
if (posEqtls != null) {
for (EQTL eQtl : posEqtls) {
if (eQtl != null && aseGenes.contains(eQtl.getProbe())) {
if (countedGenes.contains(eQtl.getProbe())) {
continue;
}
countedGenes.add(eQtl.getProbe());
// System.out.println(eQtl.getProbe());
// if(eQtl.getRsChr() == 6 && eQtl.getRsChrPos() > 20000000 && eQtl.getRsChrPos() <
// 40000000) { continue; }
++aseWithEQtl;
double aseEstimate = Double.parseDouble(elements[ASE_ESTIMATE_COLUMN]);
double eQtlZ =
elements[ASE_A1_COLUMN].equals(eQtl.getAlleleAssessed())
? eQtl.getZscore()
: eQtl.getZscore() * -1;
if (aseEstimate > 0.5 && eQtlZ > 0 || aseEstimate < 0.5 && eQtlZ < 0) {
// System.out.println("Same direction: " + eQtl.getRsChr() + ":" + eQtl.getRsChrPos()
// + "\t" + elements[ASE_A1_COLUMN] + "\t" + eQtl.getAlleleAssessed() + "\t" +
// aseEstimate + "\t" + eQtl.getZscore());
++sameDirection;
} else {
// System.out.println("Opposite: " + eQtl.getRsChr() + ":" + eQtl.getRsChrPos() + "\t"
// + elements[ASE_A1_COLUMN] + "\t" + eQtl.getAlleleAssessed() + "\t" + aseEstimate +
// "\t" + eQtl.getZscore());
++oppositeDirection;
}
}
}
}
}
NumberFormat numberFormat = NumberFormat.getInstance();
numberFormat.setMinimumFractionDigits(2);
numberFormat.setMaximumFractionDigits(2);
System.out.println("Ase total: " + aseTotal);
System.out.println(
"Ase SNP with eQTL effect: "
+ aseWithEQtl
+ " ("
+ numberFormat.format(aseWithEQtl / (double) aseTotal)
+ ")");
System.out.println(
" - Same direction: "
+ sameDirection
+ " ("
+ numberFormat.format(sameDirection / (double) aseWithEQtl)
+ ")");
System.out.println(
" - Opposite direction: "
+ oppositeDirection
+ " ("
+ numberFormat.format(oppositeDirection / (double) aseWithEQtl)
+ ")");
}
/*
* Find LD SNPs
*/
public HashMap<String, TreeMap<Integer, ArrayList<Ld>>> calculateLd(
HashMap<String, EQTL> eqtlData,
RandomAccessGenotypeData genotypeData,
int windowSize,
double r2CutOff)
throws IOException {
// Use a window size of 250k: eQTL pos - 250k and eQTL pos + 250k
Ld ld = null;
HashMap<String, TreeMap<Integer, ArrayList<Ld>>> ldResults =
new HashMap<String, TreeMap<Integer, ArrayList<Ld>>>();
Iterator<Map.Entry<String, EQTL>> eqtlIterator = eqtlData.entrySet().iterator();
while (eqtlIterator.hasNext()) {
Map.Entry pairs = (Map.Entry) eqtlIterator.next();
EQTL eqtl = (EQTL) pairs.getValue();
GeneticVariant eQtlSnp =
genotypeData.getSnpVariantByPos(eqtl.getRsChr().toString(), eqtl.getRsChrPos());
if (eQtlSnp != null) {
for (GeneticVariant gv :
genotypeData.getVariantsByRange(eqtl.getRsChr().toString(), 0, windowSize)) {
if (eQtlSnp.isBiallelic() && gv.isBiallelic()) {
try {
ld = eQtlSnp.calculateLd(gv);
} catch (LdCalculatorException ex) {
System.out.println("Error in LD calculation: " + ex.getMessage());
System.exit(1);
}
GeneticVariant variant1 = ld.getVariant1();
GeneticVariant variant2 = ld.getVariant2();
if (ld.getR2() >= r2CutOff) {
// Place results in a convenient structure for later.
TreeMap<Integer, ArrayList<Ld>> tmp;
ArrayList<Ld> ldList;
if (ldResults.containsKey(variant2.getSequenceName())) {
tmp = ldResults.get(variant2.getSequenceName());
if (tmp.containsKey(variant2.getStartPos())) {
ldList = tmp.get(variant2.getStartPos());
ldList.add(ld);
} else {
ldList = new ArrayList<Ld>();
ldList.add(ld);
tmp.put(variant2.getStartPos(), ldList);
}
} else {
tmp = new TreeMap<Integer, ArrayList<Ld>>();
ldList = new ArrayList<Ld>();
ldList.add(ld);
tmp.put(variant2.getStartPos(), ldList);
ldResults.put(variant2.getSequenceName(), tmp);
}
}
}
}
}
}
return ldResults;
}