| 1 |
/* |
| 2 |
* MIT License |
| 3 |
* |
| 4 |
* Copyright 2017 Broad Institute |
| 5 |
* |
| 6 |
* Permission is hereby granted, free of charge, to any person obtaining a copy |
| 7 |
* of this software and associated documentation files (the "Software"), to deal |
| 8 |
* in the Software without restriction, including without limitation the rights |
| 9 |
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell |
| 10 |
* copies of the Software, and to permit persons to whom the Software is |
| 11 |
* furnished to do so, subject to the following conditions: |
| 12 |
* |
| 13 |
* The above copyright notice and this permission notice shall be included in all |
| 14 |
* copies or substantial portions of the Software. |
| 15 |
* |
| 16 |
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
| 17 |
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 18 |
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
| 19 |
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
| 20 |
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 21 |
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE |
| 22 |
* SOFTWARE. |
| 23 |
*/ |
| 24 |
package org.broadinstitute.dropseqrna.barnyard.digitalallelecounts.sampleassignment.multisample; |
| 25 |
|
| 26 |
import org.apache.commons.math3.stat.StatUtils; |
| 27 |
import org.broadinstitute.dropseqrna.barnyard.digitalallelecounts.LikelihoodUtils; |
| 28 |
|
| 29 |
public class SamplePairAssignmentForCell { |
| 30 |
|
| 31 |
private final String cellBarcode; |
| 32 |
private final String sampleOne; |
| 33 |
private final String sampleTwo; |
| 34 |
|
| 35 |
// all likelihoods should be in log10 space. |
| 36 |
private final double sampleOneSingleLikelihood; |
| 37 |
private final double sampleTwoSingleLikelihood; |
| 38 |
private final double doubletLikelihood; |
| 39 |
|
| 40 |
// if a particular mixture of two cells is forced, keep that here. |
| 41 |
// private Double forcedMixture=null; |
| 42 |
// private Double forcedMixtureLikelihood=null; |
| 43 |
|
| 44 |
private final double mixture; |
| 45 |
private final int impossibleAllelesSampleOne; |
| 46 |
private final int impossibleAllelesSampleTwo; |
| 47 |
private final int numSNPs; |
| 48 |
private final int numInformativeSNPs; |
| 49 |
private final int numUMIs; |
| 50 |
private final int numInformativeUMIs; |
| 51 |
private final int numInformativeHomozygousUMIsSampleOne; |
| 52 |
private final int numInformativeHomozygousUMIsSampleTwo; |
| 53 |
|
| 54 |
// Normalize the likelihoods across many pairs of models for the same cell. |
| 55 |
// Likelihoods for each model are scaled by this factor. |
| 56 |
// This is used to select the best pair of donors out of all possible pairs, but has no other impact. |
| 57 |
private double scalingFactor; |
| 58 |
|
| 59 |
public SamplePairAssignmentForCell (final String cellBarcode, final String sampleOne, final String sampleTwo, |
| 60 |
final double sampleOneSingleLikelihood, final double sampleTwoSingleLikelihood, final double doubletLikelihood, final double mixture) { |
| 61 |
this.cellBarcode=cellBarcode; |
| 62 |
this.sampleOne=sampleOne; |
| 63 |
this.sampleTwo=sampleTwo; |
| 64 |
this.sampleOneSingleLikelihood=sampleOneSingleLikelihood; |
| 65 |
this.sampleTwoSingleLikelihood=sampleTwoSingleLikelihood; |
| 66 |
this.doubletLikelihood=doubletLikelihood; |
| 67 |
this.mixture=mixture; |
| 68 |
this.impossibleAllelesSampleOne=0; |
| 69 |
this.impossibleAllelesSampleTwo=0; |
| 70 |
this.numSNPs=0; |
| 71 |
this.numInformativeSNPs=0; |
| 72 |
this.numUMIs=0; |
| 73 |
this.numInformativeUMIs=0; |
| 74 |
this.numInformativeHomozygousUMIsSampleOne=0; |
| 75 |
this.numInformativeHomozygousUMIsSampleTwo=0; |
| 76 |
this.scalingFactor=1; |
| 77 |
} |
| 78 |
|
| 79 |
public SamplePairAssignmentForCell (final String cellBarcode, final String sampleOne, final String sampleTwo, |
| 80 |
final double sampleOneSingleLikelihood, final double sampleTwoSingleLikelihood, final double doubletLikelihood, final double mixture, |
| 81 |
final int impossibleAllelesSampleOne, final int impossibleAllelesSampleTwo, final int numInformativeSNPs, final int numSNPs, final int numUMIs, final int numInformativeUMIs, |
| 82 |
final int numInformativeHomozygousUMIsSampleOne, final int numInformativeHomozygousUMIsSampleTwo) { |
| 83 |
this.cellBarcode=cellBarcode; |
| 84 |
this.sampleOne=sampleOne; |
| 85 |
this.sampleTwo=sampleTwo; |
| 86 |
this.sampleOneSingleLikelihood=sampleOneSingleLikelihood; |
| 87 |
this.sampleTwoSingleLikelihood=sampleTwoSingleLikelihood; |
| 88 |
this.doubletLikelihood=doubletLikelihood; |
| 89 |
this.mixture=mixture; |
| 90 |
this.impossibleAllelesSampleOne=impossibleAllelesSampleOne; |
| 91 |
this.impossibleAllelesSampleTwo=impossibleAllelesSampleTwo; |
| 92 |
this.numInformativeSNPs=numInformativeSNPs; |
| 93 |
this.numSNPs=numSNPs; |
| 94 |
this.numUMIs=numUMIs; |
| 95 |
this.numInformativeUMIs=numInformativeUMIs; |
| 96 |
this.numInformativeHomozygousUMIsSampleOne=numInformativeHomozygousUMIsSampleOne; |
| 97 |
this.numInformativeHomozygousUMIsSampleTwo=numInformativeHomozygousUMIsSampleTwo; |
| 98 |
this.scalingFactor=1; |
| 99 |
} |
| 100 |
|
| 101 |
public static SamplePairAssignmentForCell constructEmptyResult (final String cellBarcode, final String sampleOne) { |
| 102 |
SamplePairAssignmentForCell r= new SamplePairAssignmentForCell(cellBarcode, sampleOne, sampleOne, 0d, -100, -101, 1, 0, 0, 0, 0, 0, 0, 0, 0); |
| 103 |
return (r); |
| 104 |
} |
| 105 |
|
| 106 |
public double getSampleOneSingleLikelihood() { |
| 107 |
return sampleOneSingleLikelihood; |
| 108 |
} |
| 109 |
|
| 110 |
public double getScaledSampleOneSingleLikelihood () { |
| 111 |
return this.sampleOneSingleLikelihood*this.scalingFactor; |
| 112 |
} |
| 113 |
|
| 114 |
public double getSampleTwoSingleLikelihood() { |
| 115 |
return sampleTwoSingleLikelihood; |
| 116 |
} |
| 117 |
|
| 118 |
public double getScaledSampleTwoSingleLikelihood() { |
| 119 |
return this.sampleTwoSingleLikelihood*this.scalingFactor; |
| 120 |
} |
| 121 |
|
| 122 |
public double getDoubletLikelihood() { |
| 123 |
return doubletLikelihood; |
| 124 |
} |
| 125 |
|
| 126 |
public double getScaledDoubletLikelihood() { |
| 127 |
return this.doubletLikelihood * this.scalingFactor; |
| 128 |
} |
| 129 |
|
| 130 |
public String getCellBarcode() { |
| 131 |
return cellBarcode; |
| 132 |
} |
| 133 |
|
| 134 |
public String getSampleOne() { |
| 135 |
return sampleOne; |
| 136 |
} |
| 137 |
|
| 138 |
public String getSampleTwo() { |
| 139 |
return sampleTwo; |
| 140 |
} |
| 141 |
|
| 142 |
public double getMixture() { |
| 143 |
return mixture; |
| 144 |
} |
| 145 |
|
| 146 |
public int getNumUMIs () { |
| 147 |
return this.numUMIs; |
| 148 |
} |
| 149 |
|
| 150 |
public int getImpossibleAllelesSampleOne() { |
| 151 |
return impossibleAllelesSampleOne; |
| 152 |
} |
| 153 |
|
| 154 |
public int getImpossibleAllelesSampleTwo() { |
| 155 |
return impossibleAllelesSampleTwo; |
| 156 |
} |
| 157 |
|
| 158 |
public int getNumInformativeHomozygousUMIsSampleOne() { |
| 159 |
return numInformativeHomozygousUMIsSampleOne; |
| 160 |
} |
| 161 |
|
| 162 |
public int getNumInformativeHomozygousUMIsSampleTwo() { |
| 163 |
return numInformativeHomozygousUMIsSampleTwo; |
| 164 |
} |
| 165 |
|
| 166 |
public int getNumSNPs() { |
| 167 |
return numSNPs; |
| 168 |
} |
| 169 |
|
| 170 |
public int getNumInformativeSNPs() { |
| 171 |
return numInformativeSNPs; |
| 172 |
} |
| 173 |
|
| 174 |
public double getDoubletLikelihoodRatio () { |
| 175 |
return this.doubletLikelihood - Math.max(this.sampleOneSingleLikelihood, this.sampleTwoSingleLikelihood); |
| 176 |
} |
| 177 |
|
| 178 |
public double getBestLikelihood () { |
| 179 |
return Math.max(this.doubletLikelihood, Math.max(this.sampleOneSingleLikelihood, this.sampleTwoSingleLikelihood)); |
| 180 |
} |
| 181 |
|
| 182 |
public double getScaledBestLikelihood () { |
| 183 |
return Math.max(getScaledDoubletLikelihood(), Math.max(getScaledSampleOneSingleLikelihood(), getScaledSampleTwoSingleLikelihood())); |
| 184 |
} |
| 185 |
|
| 186 |
/** |
| 187 |
* Probability of a doublet is the likelihood of the doublet mixture divided by the sum of all likelihoods (doublet, singlet 1, singlet 2) |
| 188 |
* @return The probability this cell is a doublet. |
| 189 |
*/ |
| 190 |
public double getDoubletPvalue () { |
| 191 |
double [] allLikelihoods = {this.sampleOneSingleLikelihood, this.sampleTwoSingleLikelihood, this.doubletLikelihood}; |
| 192 |
double maxValue = StatUtils.max(allLikelihoods); |
| 193 |
double totalLikelihood=0; |
| 194 |
|
| 195 |
for (int i=0; i<allLikelihoods.length; i++) { |
| 196 |
double d = allLikelihoods[i]; |
| 197 |
d=d-maxValue; |
| 198 |
d=Math.pow(10, d); |
| 199 |
totalLikelihood+=d; |
| 200 |
} |
| 201 |
|
| 202 |
double mixtureLike = Math.pow(10, doubletLikelihood-maxValue); |
| 203 |
double result = mixtureLike/totalLikelihood; |
| 204 |
if (result < Double.MIN_VALUE) result = Double.MIN_VALUE; |
| 205 |
|
| 206 |
return result; |
| 207 |
} |
| 208 |
|
| 209 |
|
| 210 |
public String getCombinedDonorName() { |
| 211 |
return this.sampleOne+":"+this.sampleTwo; |
| 212 |
} |
| 213 |
|
| 214 |
public String getBestSample () { |
| 215 |
if (this.doubletLikelihood> Math.max(this.sampleOneSingleLikelihood, this.sampleTwoSingleLikelihood)) |
| 216 |
return getCombinedDonorName(); |
| 217 |
else if (this.sampleOneSingleLikelihood>this.sampleTwoSingleLikelihood) |
| 218 |
return this.sampleOne; |
| 219 |
else |
| 220 |
return this.sampleTwo; |
| 221 |
} |
| 222 |
|
| 223 |
public double getScalingFactor() { |
| 224 |
return scalingFactor; |
| 225 |
} |
| 226 |
|
| 227 |
public void setScalingFactor(double scalingFactor) { |
| 228 |
this.scalingFactor = scalingFactor; |
| 229 |
} |
| 230 |
|
| 231 |
|
| 232 |
|
| 233 |
@Override |
| 234 |
public String toString () { |
| 235 |
return this.cellBarcode +" Mixture [" + this.mixture+"] like [" + this.doubletLikelihood+"] S1 [" + this.sampleOne +"="+this.sampleOneSingleLikelihood+"] S2 [" |
| 236 |
+ this.sampleTwo+"="+this.sampleTwoSingleLikelihood+"]" + " impos alleles S1 ["+ this.impossibleAllelesSampleOne +"] impos alleles S2 [" + this.impossibleAllelesSampleTwo+"] " |
| 237 |
+ "num SNPs ["+this.numSNPs+"] num informative SNPs [" + this.numInformativeSNPs +"] num UMIs [" + this.numUMIs+"] num informative UMIs [" + this.numInformativeUMIs |
| 238 |
+ "] doublet likelihood ratio test [" + getDoubletLikelihoodRatio() +"] " + "bestLikelihood [" + this.getBestLikelihood() + "] best sample [" + this.getBestSample() +"]"; |
| 239 |
} |
| 240 |
|
| 241 |
public int getNumInformativeUMIs() { |
| 242 |
return numInformativeUMIs; |
| 243 |
} |
| 244 |
|
| 245 |
|
| 246 |
@Override |
| 247 |
public int hashCode() { |
| 248 |
final int prime = 31; |
| 249 |
int result = 1; |
| 250 |
result = prime * result + ((this.cellBarcode == null) ? 0 : cellBarcode.hashCode()); |
| 251 |
long temp; |
| 252 |
temp = Double.doubleToLongBits(doubletLikelihood); |
| 253 |
result = prime * result + (int) (temp ^ (temp >>> 32)); |
| 254 |
result = prime * result + impossibleAllelesSampleOne; |
| 255 |
result = prime * result + impossibleAllelesSampleTwo; |
| 256 |
temp = Double.doubleToLongBits(mixture); |
| 257 |
result = prime * result + (int) (temp ^ (temp >>> 32)); |
| 258 |
result = prime * result + numInformativeSNPs; |
| 259 |
result = prime * result + numInformativeUMIs; |
| 260 |
result = prime * result + numSNPs; |
| 261 |
result = prime * result + numUMIs; |
| 262 |
result = prime * result + ((sampleOne == null) ? 0 : sampleOne.hashCode()); |
| 263 |
temp = Double.doubleToLongBits(sampleOneSingleLikelihood); |
| 264 |
result = prime * result + (int) (temp ^ (temp >>> 32)); |
| 265 |
result = prime * result + ((sampleTwo == null) ? 0 : sampleTwo.hashCode()); |
| 266 |
temp = Double.doubleToLongBits(sampleTwoSingleLikelihood); |
| 267 |
result = prime * result + (int) (temp ^ (temp >>> 32)); |
| 268 |
return result; |
| 269 |
} |
| 270 |
|
| 271 |
@Override |
| 272 |
public boolean equals(Object obj) { |
| 273 |
if (this == obj) |
| 274 |
return true; |
| 275 |
if (obj == null) |
| 276 |
return false; |
| 277 |
if (getClass() != obj.getClass()) |
| 278 |
return false; |
| 279 |
SamplePairAssignmentForCell other = (SamplePairAssignmentForCell) obj; |
| 280 |
if (cellBarcode == null) { |
| 281 |
if (other.cellBarcode != null) |
| 282 |
return false; |
| 283 |
} else if (!cellBarcode.equals(other.cellBarcode)) |
| 284 |
return false; |
| 285 |
if (Double.doubleToLongBits(doubletLikelihood) != Double.doubleToLongBits(other.doubletLikelihood)) |
| 286 |
return false; |
| 287 |
if (impossibleAllelesSampleOne != other.impossibleAllelesSampleOne) |
| 288 |
return false; |
| 289 |
if (impossibleAllelesSampleTwo != other.impossibleAllelesSampleTwo) |
| 290 |
return false; |
| 291 |
if (Double.doubleToLongBits(mixture) != Double.doubleToLongBits(other.mixture)) |
| 292 |
return false; |
| 293 |
if (numInformativeSNPs != other.numInformativeSNPs) |
| 294 |
return false; |
| 295 |
if (numInformativeUMIs != other.numInformativeUMIs) |
| 296 |
return false; |
| 297 |
if (numSNPs != other.numSNPs) |
| 298 |
return false; |
| 299 |
if (numUMIs != other.numUMIs) |
| 300 |
return false; |
| 301 |
if (sampleOne == null) { |
| 302 |
if (other.sampleOne != null) |
| 303 |
return false; |
| 304 |
} else if (!sampleOne.equals(other.sampleOne)) |
| 305 |
return false; |
| 306 |
if (Double.doubleToLongBits(sampleOneSingleLikelihood) != Double |
| 307 |
.doubleToLongBits(other.sampleOneSingleLikelihood)) |
| 308 |
return false; |
| 309 |
if (sampleTwo == null) { |
| 310 |
if (other.sampleTwo != null) |
| 311 |
return false; |
| 312 |
} else if (!sampleTwo.equals(other.sampleTwo)) |
| 313 |
return false; |
| 314 |
if (Double.doubleToLongBits(sampleTwoSingleLikelihood) != Double |
| 315 |
.doubleToLongBits(other.sampleTwoSingleLikelihood)) |
| 316 |
return false; |
| 317 |
return true; |
| 318 |
} |
| 319 |
|
| 320 |
|
| 321 |
/* |
| 322 |
public Double getForcedMixture() { |
| 323 |
return forcedMixture; |
| 324 |
} |
| 325 |
|
| 326 |
public Double getForcedMixtureLikelihood() { |
| 327 |
return forcedMixtureLikelihood; |
| 328 |
} |
| 329 |
*/ |
| 330 |
|
| 331 |
/** |
| 332 |
* Compares the doublet likelihood to the two single likelihoods. |
| 333 |
* If the doublet likelihood is equal to the sampleOneSingleLikelihood, then |
| 334 |
* @param originalMixture |
| 335 |
* @param doubletLikelihood |
| 336 |
* @param sampleOneSingleLikelihood |
| 337 |
* @param sampleTwoSingleLikelihood |
| 338 |
*/ |
| 339 |
/* |
| 340 |
static double getMixture (final double originalMixture, final double doubletLikelihood, final double sampleOneSingleLikelihood, final double sampleTwoSingleLikelihood) { |
| 341 |
double epsilon = Math.min(doubletLikelihood,sampleOneSingleLikelihood)*0.01; |
| 342 |
int cmp = Precision.compareTo(doubletLikelihood, sampleOneSingleLikelihood, epsilon); |
| 343 |
// the first sample likelihood explains the data as well as the doublet likelihood. The mixture is 1. |
| 344 |
if (cmp==0) return 1; |
| 345 |
|
| 346 |
epsilon = Math.min(doubletLikelihood,sampleTwoSingleLikelihood)*0.01; |
| 347 |
cmp = Precision.compareTo(doubletLikelihood, sampleTwoSingleLikelihood, epsilon); |
| 348 |
// the second sample likelihood explains the data as well as the doublet likelihood. The mixture is 0. |
| 349 |
if (cmp==0) return 0; |
| 350 |
|
| 351 |
return originalMixture; |
| 352 |
|
| 353 |
} |
| 354 |
*/ |
| 355 |
|
| 356 |
|
| 357 |
} |