1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376
|
#! /usr/bin/env python
# Copyright 2011 Martin C. Frith
# This script reads alignments of DNA reads to a genome, and estimates
# the probability that each alignment represents the genomic source of
# the read. It assumes that the reads come in pairs, where each pair
# is from either end of a DNA fragment.
# A write-up of the method is available from the author (and should be
# published somewhere).
# Seems to work with Python 2.x, x>=4.
# The --rna option makes it assume that the genomic fragment lengths
# follow a log-normal distribution (instead of a normal distribution).
# In one test with human RNA, log-normal was a remarkably good fit,
# but not perfect. The true distribution looked like a mixture of 2
# log-normals: a dominant one for shorter introns, and a minor one for
# huge introns. Thus, our use of a single log-normal fails to model
# rare, huge introns. To compensate for that, the default value of
# --disjoint is increased when --rna is used.
# (Should we try to estimate the prior probability of disjoint mapping
# from the data? But maybe ignore low-scoring alignments for that?
# Estimate disjoint maps to opposite strands of same chromosome = maps
# to same strand of same chromosome?)
import itertools, math, operator, optparse, os, signal, sys
def logSumExp(numbers):
"""Adds numbers, in log space, to avoid overflow."""
n = list(numbers)
if not n: return -1e99 # should be -inf
m = max(n)
s = sum(math.exp(i - m) for i in n) # fsum is only Python >= 2.6.
return math.log(s) + m
def warn(*things):
prog = os.path.basename(sys.argv[0])
text = " ".join(map(str, things))
sys.stderr.write(prog + ": " + text + "\n")
# Kludge: using "~" to indicate "finished".
def safeNext(groupedby):
try:
k, g = groupedby.next()
return k, list(g)
except StopIteration:
return "~", []
def joinby(iterable1, iterable2, keyfunc):
"""Yields groups from iterable1 and iterable2 that share the same key."""
groups1 = itertools.groupby(iterable1, keyfunc)
groups2 = itertools.groupby(iterable2, keyfunc)
k1, v1 = safeNext(groups1)
k2, v2 = safeNext(groups2)
while k1 != "~" or k2 != "~":
if k1 < k2:
yield v1, []
k1, v1 = safeNext(groups1)
elif k1 > k2:
yield [], v2
k2, v2 = safeNext(groups2)
else:
yield v1, v2
k1, v1 = safeNext(groups1)
k2, v2 = safeNext(groups2)
class Alignment:
def __init__(self, pairName, chromName, strand, endpoint, chromSize,
score, lines):
self.pairName = pairName
self.chromName = chromName
self.strand = strand
self.endpoint = endpoint
self.chromSize = chromSize
self.score = score
self.lines = lines
class AlignmentParameters:
"""Parses the t (temperature) and e (minimum score) parameters."""
def __init__(self): # dummy values:
self.t = -1
self.e = -1
def update(self, line):
for i in line.split():
if self.t == -1 and i.startswith("t="):
self.t = float(i[2:])
if self.t <= 0: raise Exception("t must be positive")
if self.e == -1 and i.startswith("e="):
self.e = float(i[2:])
if self.e <= 0: raise Exception("e must be positive")
def validate(self):
if self.t == -1: raise Exception("I need a header line with t=")
if self.e == -1: raise Exception("I need a header line with e=")
def printAlignmentWithMismapProb(lines, prob):
p = "%.3g" % prob
if len(lines) == 1: # we have tabular format
print lines[0].rstrip() + "\t" + p
else: # we have MAF format
print lines[0].rstrip() + " mismap=" + p
for i in lines[1:]: print i,
def fragmentLength(alignment1, alignment2):
length = alignment1.endpoint + alignment2.endpoint
if length > alignment1.chromSize: length -= alignment1.chromSize
return length
def conjointScores(aln1, alns2, opts): # maybe slow
for i in alns2:
if i.chromName != aln1.chromName or i.strand == aln1.strand: continue
length = fragmentLength(aln1, i)
if length <= 0: continue
if opts.rna: # use a log-normal distribution
loglen = math.log(length)
yield i.score + opts.inner * (loglen - opts.fraglen) ** 2 - loglen
else: # use a normal distribution
yield i.score + opts.inner * (length - opts.fraglen) ** 2
def printAlignmentsForOneRead(alignments1, alignments2, opts, maxMissingScore):
if alignments2:
x = opts.disjointScore + logSumExp(i.score for i in alignments2)
for i in alignments1:
y = opts.outer + logSumExp(conjointScores(i, alignments2, opts))
i.z = i.score + logSumExp((x, y))
w = maxMissingScore + max(i.score for i in alignments2)
else:
for i in alignments1:
i.z = i.score + opts.disjointScore
w = maxMissingScore
z = logSumExp(i.z for i in alignments1)
zw = logSumExp((z, w))
for i in alignments1:
prob = 1 - math.exp(i.z - zw)
if prob <= opts.mismap: printAlignmentWithMismapProb(i.lines, prob)
def measurablePairs(alignments1, alignments2): # maybe slow
"""Yields alignment pairs on opposite strands of the same chromosome."""
for i in alignments1:
for j in alignments2:
if i.chromName == j.chromName and i.strand != j.strand:
yield i, j
def unambiguousFragmentLength(alignments1, alignments2):
"""Returns the fragment length implied by alignments of a pair of reads."""
old = None
for i, j in measurablePairs(alignments1, alignments2):
new = fragmentLength(i, j)
if old is None: old = new
elif new != old: return None # the fragment length is ambiguous
return old
def unambiguousFragmentLengths(queryPairs):
for i in queryPairs:
length = unambiguousFragmentLength(*i)
if length is not None: yield length
def checkChromSize(chromSizes, rName, rSize):
if rName in chromSizes and chromSizes[rName] != rSize:
raise Exception("inconsistent lengths for: " + rName)
chromSizes[rName] = rSize
def parseMafScore(aLine):
for i in aLine.split():
if i.startswith("score="): return i[6:]
raise Exception("missing score")
def parseMaf(lines):
aLines = [i for i in lines if i[0] == "a"]
sLines = [i for i in lines if i[0] == "s"]
score = parseMafScore(aLines[0])
rWords = sLines[0].split()[1:6]
qWords = sLines[1].split()[1:6]
return score, rWords, qWords, lines
def parseTab(line):
words = line.split()
score = words[0]
rWords = words[1:6]
qWords = words[6:11]
return score, rWords, qWords, [line]
def readAlignmentData(lines, params):
"""Yields alignment data from MAF or tabular format, and updates params."""
mafLines = []
for line in lines:
if line[0] == "#":
params.update(line)
elif line[0].isdigit():
yield parseTab(line)
elif line.isspace():
if mafLines:
mafLines.append(line)
yield parseMaf(mafLines)
mafLines = []
else:
mafLines.append(line)
if mafLines:
yield parseMaf(mafLines)
def cunningCoordinate(strand, rStart, rSpan, rSize, isCircular):
"""Returns a coordinate, that allows fast calculation of frag lengths."""
if strand == "+": return -rStart
elif isCircular: return rStart + rSpan + rSize
else: return rStart + rSpan
def infoFromAlignmentWords(words):
seqName, alnStart, alnSize, strand, seqSize = words
return seqName, int(alnStart), int(alnSize), strand, int(seqSize)
def readAlignments(lines, chromSizes, params, circularChroms):
"""Yields alignments, checks their order, updates chromSizes and params."""
oldName = ""
for i in readAlignmentData(lines, params):
score, rWords, qWords, text = i
rName, rStart, rSpan, rStrand, rSize = infoFromAlignmentWords(rWords)
qName, qStart, qSpan, qStrand, qSize = infoFromAlignmentWords(qWords)
index = qName.rfind("/")
if index < 0: pairName = qName
else: pairName = qName[:index+1]
if pairName < oldName:
raise Exception("alignments not sorted properly")
oldName = pairName
isCircular = rName in circularChroms or "." in circularChroms
c = cunningCoordinate(qStrand, rStart, rSpan, rSize, isCircular)
checkChromSize(chromSizes, rName, rSize) # needed in 1st pass
scaledScore = float(score) / params.t # needed in 2nd pass
yield Alignment(pairName, rName, qStrand, c, rSize, scaledScore, text)
def estimateFragmentLengthDistribution(lengths, opts):
if not lengths:
raise Exception("can't estimate fragment length distribution")
# Define quartiles in the most naive way possible:
lengths.sort()
sampleSize = len(lengths)
quartile1 = lengths[sampleSize // 4]
quartile2 = lengths[sampleSize // 2]
quartile3 = lengths[sampleSize * 3 // 4]
warn("fragment length sample size:", sampleSize)
warn("fragment length quartiles:", quartile1, quartile2, quartile3)
if opts.fraglen is None:
if opts.rna: opts.fraglen = math.log(quartile2)
else: opts.fraglen = quartile2
warn("estimated mean fragment length:", opts.fraglen)
if opts.sdev is None:
if opts.rna: iqr = math.log(quartile3) - math.log(quartile1)
else: iqr = quartile3 - quartile1
opts.sdev = iqr / 1.35 # Normal Distribution
warn("estimated standard deviation:", "%g" % opts.sdev)
if quartile1 <= 0:
raise Exception("too many fragment lengths <= 0")
def safeLog(x):
if x == 0: return -1e99
else: return math.log(x)
def calculateScorePieces(opts, params1, params2):
if opts.sdev == 0:
if opts.rna: opts.outer = opts.fraglen
else: opts.outer = 0
opts.inner = -1e99
else: # parameters for a Normal Distribution (of fragment lengths):
opts.outer = -math.log(opts.sdev * math.sqrt(2 * math.pi))
opts.inner = -1.0 / (2 * opts.sdev ** 2)
opts.outer += safeLog(1 - opts.disjoint)
# Multiply genome size by 2, because it has 2 strands:
opts.disjointScore = safeLog(opts.disjoint) - math.log(opts.genome * 2)
#opts.disjointScore = math.log(opts.disjoint / (opts.genome * 2.0))
# Max possible influence of an alignment just below the score threshold:
maxLogPrior = opts.outer
if opts.rna: maxLogPrior += opts.sdev ** 2 / 2 - opts.fraglen
opts.maxMissingScore1 = (params1.e - 1) / params1.t + maxLogPrior
opts.maxMissingScore2 = (params2.e - 1) / params2.t + maxLogPrior
def lastPairProbs(opts, args):
fileName1, fileName2 = args
params1 = AlignmentParameters()
params2 = AlignmentParameters()
chromSizes = {}
in1 = open(fileName1)
in2 = open(fileName2)
alns1 = readAlignments(in1, chromSizes, params1, opts.circular)
alns2 = readAlignments(in2, chromSizes, params2, opts.circular)
queryPairs = joinby(alns1, alns2, operator.attrgetter("pairName"))
lengths = list(unambiguousFragmentLengths(queryPairs))
in1.close()
in2.close()
params1.validate()
params2.validate()
if opts.fraglen is None or opts.sdev is None:
estimateFragmentLengthDistribution(lengths, opts)
if opts.sdev < 0: raise Exception("standard deviation < 0")
if opts.genome is None:
opts.genome = sum(chromSizes.values())
warn("genome size:", opts.genome)
if opts.genome <= 0: raise Exception("genome size <= 0")
calculateScorePieces(opts, params1, params2)
in1 = open(fileName1)
in2 = open(fileName2)
alns1 = readAlignments(in1, chromSizes, params1, opts.circular)
alns2 = readAlignments(in2, chromSizes, params2, opts.circular)
queryPairs = joinby(alns1, alns2, operator.attrgetter("pairName"))
for i, j in queryPairs:
printAlignmentsForOneRead(i, j, opts, opts.maxMissingScore1)
printAlignmentsForOneRead(j, i, opts, opts.maxMissingScore2)
in1.close()
in2.close()
if __name__ == "__main__":
signal.signal(signal.SIGPIPE, signal.SIG_DFL) # avoid silly error message
usage = """
%prog --help
%prog [options] alignments1 alignments2"""
description = "Read alignments of paired DNA reads to a genome, and estimate the probability that each alignment represents the genomic source of the read."
op = optparse.OptionParser(usage=usage, description=description)
op.add_option("-r", "--rna", action="store_true", help=
"specifies that the reads are from potentially-spliced RNA")
op.add_option("-m", "--mismap", type="float", default=0.01, metavar="M",
help="don't write alignment pairs with mismap probability > M (default: %default)")
op.add_option("-f", "--fraglen", type="float", metavar="BP",
help="mean fragment length in bp")
op.add_option("-s", "--sdev", type="float", metavar="BP",
help="standard deviation of fragment length")
op.add_option("-g", "--genome", type="float", metavar="BP",
help="haploid genome size in bp")
op.add_option("-d", "--disjoint", type="float",
metavar="PROB", help=
"prior probability of disjoint mapping (default: 0.02 if -r, else 0.01)")
op.add_option("-c", "--circular", action="append", metavar="CHROM",
help="specifies that chromosome CHROM is circular (default: chrM)")
(opts, args) = op.parse_args()
if opts.disjoint is None:
if opts.rna: opts.disjoint = 0.02
else: opts.disjoint = 0.01
if opts.disjoint < 0: op.error("option -d: should be >= 0")
if opts.disjoint > 1: op.error("option -d: should be <= 1")
if len(args) != 2: op.error("please give me two file names")
if opts.circular is None: opts.circular = ["chrM"]
try: lastPairProbs(opts, args)
except KeyboardInterrupt: pass # avoid silly error message
except Exception, e:
warn("error:", e)
sys.exit(1)
|