private double calcPval(
HashMap<Integer, String> singleGeneCaseValueMap, String profileType, String profileStableId)
throws IllegalArgumentException {
double _p_value = 0.0;
if (profileType.equals(GeneticAlterationType.MUTATION_EXTENDED.toString())
|| profileType.equals(GeneticAlterationType.COPY_NUMBER_ALTERATION.toString())) {
_p_value = runFisherExactTest(singleGeneCaseValueMap, profileType);
} else if (profileType.equals(GeneticAlterationType.MRNA_EXPRESSION.toString())
|| profileType.equals(GeneticAlterationType.PROTEIN_LEVEL.toString())) {
_p_value = runTTest(singleGeneCaseValueMap, profileStableId);
}
return _p_value;
}
private double runFisherExactTest(
HashMap<Integer, String> singleGeneCaseValueMap, String profileType) {
int a = 0, // non altered
b = 0, // x non altered, y altered
c = 0, // x altered, y non altered
d = 0; // both alered
for (Integer alteredSampleId : alteredSampleIds) {
if (singleGeneCaseValueMap.containsKey(alteredSampleId)) {
if (profileType.equals(GeneticAlterationType.COPY_NUMBER_ALTERATION.toString())) {
try {
Double value = Double.parseDouble(singleGeneCaseValueMap.get(alteredSampleId));
if (copyNumType.equals("del")) {
if (value == -2.0) {
d += 1;
} else {
c += 1;
}
} else if (copyNumType.equals("amp")) {
if (value == 2.0) {
d += 1;
} else {
c += 1;
}
}
} catch (NumberFormatException e) {
e.getStackTrace();
}
} else if (profileType.equals(GeneticAlterationType.MUTATION_EXTENDED.toString())) {
String value = singleGeneCaseValueMap.get(alteredSampleId);
if (value.equals("Non")) {
c += 1;
} else {
d += 1;
}
}
}
}
for (Integer unalteredSampleId : unalteredSampleIds) {
if (singleGeneCaseValueMap.containsKey(unalteredSampleId)) {
if (profileType.equals(GeneticAlterationType.COPY_NUMBER_ALTERATION.toString())) {
try {
Double value = Double.parseDouble(singleGeneCaseValueMap.get(unalteredSampleId));
if (copyNumType.equals("del")) {
if (value == -2.0) {
b += 1;
} else {
a += 1;
}
} else if (copyNumType.equals("amp")) {
if (value == 2.0) {
b += 1;
} else {
a += 1;
}
}
} catch (NumberFormatException e) {
e.getStackTrace();
}
} else if (profileType.equals(GeneticAlterationType.MUTATION_EXTENDED.toString())) {
String value = singleGeneCaseValueMap.get(unalteredSampleId);
if (value.equals("Non")) {
a += 1;
} else {
b += 1;
}
}
}
}
FisherExact fisher = new FisherExact(a + b + c + d);
return fisher.getCumlativeP(a, b, c, d);
}
private double calcPct(
HashMap<Integer, String> singleGeneCaseValueMap,
String profileType,
String groupType) { // group type: altered or unaltered
double _result_pct = 0, _count = 0; // altered samples count
if (profileType.equals(GeneticAlterationType.COPY_NUMBER_ALTERATION.toString())
&& copyNumType.equals("del")) {
switch (groupType) {
case "altered":
for (Integer alteredSampleId : alteredSampleIds) {
if (singleGeneCaseValueMap.containsKey(alteredSampleId)) {
try {
if (Double.parseDouble(singleGeneCaseValueMap.get(alteredSampleId)) == -2.0) {
_count += 1;
}
} catch (NumberFormatException e) {
e.getStackTrace();
}
}
}
_result_pct = _count / alteredSampleIds.size();
break;
case "unaltered":
for (Integer unalteredSampleId : unalteredSampleIds) {
if (singleGeneCaseValueMap.containsKey(unalteredSampleId)) {
try {
if (Double.parseDouble(singleGeneCaseValueMap.get(unalteredSampleId)) == -2.0) {
_count += 1;
}
} catch (NumberFormatException e) {
e.getStackTrace();
}
}
}
_result_pct = _count / unalteredSampleIds.size();
break;
}
} else if (profileType.equals(GeneticAlterationType.COPY_NUMBER_ALTERATION.toString())
&& copyNumType.equals("amp")) {
switch (groupType) {
case "altered":
for (Integer alteredSampleId : alteredSampleIds) {
if (singleGeneCaseValueMap.containsKey(alteredSampleId)) {
try {
if (Double.parseDouble(singleGeneCaseValueMap.get(alteredSampleId)) == 2.0) {
_count += 1;
}
} catch (NumberFormatException e) {
e.getStackTrace();
}
}
}
_result_pct = _count / alteredSampleIds.size();
break;
case "unaltered":
for (Integer unalteredSampleId : unalteredSampleIds) {
if (singleGeneCaseValueMap.containsKey(unalteredSampleId)) {
try {
if (Double.parseDouble(singleGeneCaseValueMap.get(unalteredSampleId)) == 2.0) {
_count += 1;
}
} catch (NumberFormatException e) {
e.getStackTrace();
}
}
}
_result_pct = _count / unalteredSampleIds.size();
break;
}
} else if (profileType.equals(GeneticAlterationType.MUTATION_EXTENDED.toString())) {
switch (groupType) {
case "altered":
for (Integer alteredSampleId : alteredSampleIds) {
if (singleGeneCaseValueMap.containsKey(alteredSampleId)) {
if (!singleGeneCaseValueMap.get(alteredSampleId).equals("Non")) {
_count += 1;
}
}
}
_result_pct = _count / alteredSampleIds.size();
break;
case "unaltered":
for (Integer unalteredSampleId : unalteredSampleIds) {
if (singleGeneCaseValueMap.containsKey(unalteredSampleId)) {
if (!singleGeneCaseValueMap.get(unalteredSampleId).equals("Non")) {
_count += 1;
}
}
}
_result_pct = _count / unalteredSampleIds.size();
break;
}
}
return _result_pct;
}
@Override
public ObjectNode process(long entrezGeneId, String[] values, ArrayList<Integer> sampleList) {
ObjectNode _datum = mapper.createObjectNode();
// create map to pair sample and value
HashMap<Integer, String> mapSampleValue = new HashMap<>();
for (int i = 0; i < values.length; i++) {
String value = values[i];
Integer sampleId = sampleList.get(i);
mapSampleValue.put(sampleId, value);
}
// remove empty entry
Iterator it = mapSampleValue.entrySet().iterator();
while (it.hasNext()) {
Map.Entry pair = (Map.Entry) it.next();
if (pair.getValue().equals("NA")
|| pair.getValue().equals("NaN")
|| pair.getValue().equals("null")) {
it.remove();
}
}
// get Gene Name and Cytoband
DaoGeneOptimized daoGeneOptimized = DaoGeneOptimized.getInstance();
String geneName = daoGeneOptimized.getGene(entrezGeneId).getHugoGeneSymbolAllCaps();
String cytoband = daoGeneOptimized.getGene(entrezGeneId).getCytoband();
// statistics analysis
if (profileType.equals(GeneticAlterationType.COPY_NUMBER_ALTERATION.toString())) {
if (!(Arrays.asList(queriedGenes)).contains(geneName)) { // remove queried genes from result
_datum.put(COL_NAME_GENE, geneName);
_datum.put(COL_NAME_CYTOBAND, cytoband);
_datum.put(
COL_NAME_PCT_ALTERED,
Integer.toString(countAltered(mapSampleValue, profileType, "altered"))
+ "////"
+ Double.toString(calcPct(mapSampleValue, profileType, "altered")));
_datum.put(
COL_NAME_PCT_UNALTERED,
Integer.toString(countAltered(mapSampleValue, profileType, "unaltered"))
+ "////"
+ Double.toString(calcPct(mapSampleValue, profileType, "unaltered")));
_datum.put(
COL_NAME_RATIO,
calcRatio(
calcPct(mapSampleValue, profileType, "altered"),
calcPct(mapSampleValue, profileType, "unaltered")));
_datum.put(COL_NAME_DIRECTION, "place holder"); // calculation is done by the front-end
_datum.put(COL_NAME_P_VALUE, calcPval(mapSampleValue, profileType, geneticProfileStableId));
if (!(calcPct(mapSampleValue, profileType, "altered") == 0.0
&& calcPct(mapSampleValue, profileType, "unaltered") == 0.0)
&& !Double.isNaN(calcPval(mapSampleValue, profileType, geneticProfileStableId))) {
return _datum;
}
}
} else if (profileType.equals(GeneticAlterationType.MRNA_EXPRESSION.toString())) {
_datum.put(COL_NAME_GENE, geneName);
_datum.put(COL_NAME_CYTOBAND, cytoband);
_datum.put(
COL_NAME_MEAN_ALTERED, calcMean(mapSampleValue, "altered", geneticProfileStableId));
_datum.put(
COL_NAME_MEAN_UNALTERED, calcMean(mapSampleValue, "unaltered", geneticProfileStableId));
_datum.put(
COL_NAME_STDEV_ALTERED, calcSTDev(mapSampleValue, "altered", geneticProfileStableId));
_datum.put(
COL_NAME_STDEV_UNALTERED, calcSTDev(mapSampleValue, "unaltered", geneticProfileStableId));
_datum.put(COL_NAME_P_VALUE, calcPval(mapSampleValue, profileType, geneticProfileStableId));
if (!Double.isNaN(calcPval(mapSampleValue, profileType, geneticProfileStableId))) {
return _datum;
}
} else if (profileType.equals(GeneticAlterationType.PROTEIN_LEVEL.toString())) {
_datum.put(COL_NAME_GENE, geneName);
_datum.put(COL_NAME_CYTOBAND, cytoband);
_datum.put(
COL_NAME_MEAN_ALTERED, calcMean(mapSampleValue, "altered", geneticProfileStableId));
_datum.put(
COL_NAME_MEAN_UNALTERED, calcMean(mapSampleValue, "unaltered", geneticProfileStableId));
_datum.put(
COL_NAME_STDEV_ALTERED, calcSTDev(mapSampleValue, "altered", geneticProfileStableId));
_datum.put(
COL_NAME_STDEV_UNALTERED, calcSTDev(mapSampleValue, "unaltered", geneticProfileStableId));
_datum.put(COL_NAME_P_VALUE, calcPval(mapSampleValue, profileType, geneticProfileStableId));
if (!Double.isNaN(calcPval(mapSampleValue, profileType, geneticProfileStableId))) {
return _datum;
}
}
return null;
}
