forked from EichlerLab/chm1_scripts
-
Notifications
You must be signed in to change notification settings - Fork 0
/
DotPlot.py.bak
314 lines (264 loc) · 10.7 KB
/
DotPlot.py.bak
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
#!/usr/bin/env python
import pandas
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import re
from Bio import SeqIO
from Bio import SeqRecord
from Bio import Seq
import argparse
import matplotlib.lines as mlines
import math
import numpy as np
import tempfile
import subprocess
import os
import Tools
import pdb
def IndexSeq(seq, k):
index = {}
for i in range(len(seq)- k + 1):
kmer = seq[i:i+k]
if (kmer not in index):
index[kmer] = []
index[kmer].append(i)
return index
def DotDirectional(query, targetIndex, k, d):
hits = []
for i in range(len(query) - k + 1):
word = query[i:i+k]
if (word in targetIndex):
for pos in targetIndex[word]:
hits.append((i, pos, d))
return hits
def DotPlot(query, target, k):
targetIndex = IndexSeq(target.upper(),k)
forward = DotDirectional(query, targetIndex, k, 0)
revTarget = Seq.Seq(target).reverse_complement().tostring().upper()
revTargetIndex= IndexSeq(revTarget, k)
reverse = DotDirectional(query, revTargetIndex, k, 1)
return forward + reverse
queryCoordRe = re.compile('(.*)_(\d+)_(\d+)/.*')
ap = argparse.ArgumentParser(description = "make a dot plot")
ap.add_argument("--query", help="Read the query sequence. Specify a read using read:bas_fil_name.bas.h5:hole_number", required=True)
ap.add_argument("--target", help="Read the target sequence.", required=True)
ap.add_argument("--matches", help="Read the matches (m1 format), or dot:k to do a simple dot plot", required=False, default=None)
ap.add_argument("--savefig", help="Write figure to a file.", default=None)
ap.add_argument("--region", help="Specify a region in the format ref:start-end .",default=None)
ap.add_argument("--blpath", help="Path to alternate blasr.", default="")
args = ap.parse_args()
tmpFiles = []
# set up query information
if (args.query[0:4] == "bas:"):
# The input is a read, not a fasta file, extract that for the run.
readValues = args.query.split(':')
basName = readValues[1]
holeNumber = int(readValues[2])
tmpQueryName = tempfile.mktemp(suffix=".query.fasta", dir=".")
tmpQuery = open(tmpQueryName, 'w')
tmpFiles.append(tmpQueryName)
command = "/net/eichler/vol5/home/mchaisso/software/blasr_1/cpp/pbihdfutils/bin/pls2fasta {} {} -holeNumber {} -trimByRegion".format(basName, tmpQueryName, holeNumber)
print command
args.query = tmpQueryName
print "extracing read to " + tmpQueryName
subprocess.call(command.split())
# set up taget
if (args.region is not None):
faiFileName = args.target + ".fai"
if (os.path.exists(faiFileName) == False):
print "Error, when using a region there must be an associated .fai file."
sys.exit(1)
targetfd = file(args.target, 'r')
fai = Tools.ReadFAIFile(faiFileName)
seq = Tools.ExtractSeq(Tools.ParseRegionStr(args.region), targetfd, fai)
tmpTargetName = tempfile.mktemp(suffix=".target.fasta", dir=".")
tmpTarget = open(tmpTargetName, 'w')
tmpFiles.append(tmpTargetName)
record = SeqRecord.SeqRecord(Seq.Seq(seq), name="",id=args.region,description="")
SeqIO.write(record, tmpTarget, "fasta")
targetName = tmpTarget.name
tmpTarget.close()
args.target = targetName
if (args.matches == None):
# it is necessary to generate the matches
targetName = args.target
matchFileName = tempfile.mktemp(suffix=".m4", dir=".")
matchFile = open(matchFileName, 'w')
tmpFiles.append(matchFileName)
args.matches = matchFileName
blasrExe = args.blpath + "blasr"
command = "{} {} {} -m 1 -out {} -bestn 100 -noPrintSubreadTitle ".format(blasrExe, args.query, args.target, args.matches)
print command
subprocess.call(command.split())
elif (args.matches[0:4] == "dot:"):
vals = args.matches.split(":")
k = int(vals[1])
queryFile = open(args.query, 'r')
targetFile = open(args.target, 'r')
matchFileName = tempfile.mktemp(suffix=".m4", dir=".")
matchFile = open(matchFileName, 'w')
tmpFiles.append(matchFileName)
args.matches = matchFileName
for query in SeqIO.parse(queryFile, "fasta"):
for target in SeqIO.parse(targetFile, "fasta"):
matches = DotPlot(query.seq.tostring(), target.seq.tostring(), k)
lenQuery = len(query)
lenTarget = len(target)
print "writing " + str(len(matches) ) + " matches "
for match in matches:
# 0 1 2 3 4 5 6 7 8 9 10 11 12
# scf7180000000013_1000_2000/0_1000 scf7180000000013 0 0 -5000 100 1000 2000 11963 0 1000 1000 20968
matchFile.write("{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\n".format(query.id, target.id, 0, match[2], -100, 100, match[1], match[1]+k, lenTarget, match[0], match[0]+k, lenQuery, 1))
matchFile.close()
qf = open(args.query, 'r')
tf = open(args.target, 'r')
mf = open(args.matches, 'r')
query = list(SeqIO.parse(qf, "fasta"))
target = list(SeqIO.parse(tf, "fasta"))
queryDict = {q.name: len(q.seq) for q in query }
targetDict = {t.name: len(t.seq) for t in target }
queryNames = [q.name for q in query]
targetNames = [t.name for t in target]
numTargets = len(targetDict)
queryLengths = [len(q.seq) for q in query]
targetLengths = [len(t.seq) for t in target]
targetCumSum = [0]*(len(target)+1)
for i in range(len(target)):
targetCumSum[i+1] = targetCumSum[i] + len(target[i].seq)
queryCumSum = [0]*(len(query) + 1)
for i in range(len(query)):
queryCumSum[i+1] = queryCumSum[i] + len(query[i].seq)
tFrac = [float(targetCumSum[i])/targetCumSum[-1] for i in range(len(targetCumSum)-1)]
qFrac = [float(queryCumSum[i])/queryCumSum[-1] for i in range(len(queryCumSum)-1)]
tStep = step=max(1,math.pow(10,math.floor(math.log10(targetCumSum[-1]))-1))
qStep = step=max(1,math.pow(10,math.floor(math.log10(queryCumSum[-1]))-1))
tTicks = [ np.arange(0, t, step=tStep) for t in targetLengths ]
qTicks = [ np.arange(0, q, step=qStep) for q in queryLengths]
tTickPositions = [ tTicks[i] + targetCumSum[i] for i in range(len(targetLengths))]
qTickPositions = [ qTicks[i] + queryCumSum[i] for i in range(len(queryLengths))]
ntTicks = sum(len(t) for t in tTicks)
nqTicks = sum(len(q) for q in qTicks)
def SetTickLists(raggedTicks, raggedTickOffset):
nticks = sum(len(t) for t in raggedTicks)
tickLayoutPos = [0]*nticks
tickLayoutText = [""] *nticks
idx = 0
for i in range(len(raggedTicks)):
for j in range(len(raggedTicks[i])):
tickLayoutPos[idx] = raggedTickOffset[i][j]
tickLayoutText[idx] = str(int(raggedTicks[i][j]))
idx+=1
return tickLayoutPos, tickLayoutText
tLayoutPos, tLayoutText = SetTickLists(tTicks, tTickPositions)
qLayoutPos, qLayoutText = SetTickLists(qTicks, qTickPositions)
queryi = { queryNames[i] : i for i in range(len(queryNames))}
targeti = {targetNames[i] : i for i in range(len(targetNames))}
querycsd = { queryNames[i] : queryCumSum[i] for i in range(len(queryNames))}
targetcsd = { targetNames[i] : targetCumSum[i] for i in range(len(targetNames))}
class Match:
def __init__(self, qname, qfragstart, qfragend, qstrand, qstart, qend, qlen, tname, tstrand, tstart, tend, tlen, pctident):
self.qname = qname
self.qfragstart = qfragstart
self.qfragend = qfragend
self.qstrand = qstrand
self.qstart = qstart
self.qend = qend
self.qlen = qlen
self.tname = tname
self.tstrand = tstrand
self.tstart = tstart
self.tend = tend
self.tlen = tlen
self.pctident = pctident
def ParseMatchLine(ml):
mv = ml.split()
# parse this format
# 0 1 2 3 4 5 6 7 8 9 10 11 12
# scf7180000000013_1000_2000/0_1000 scf7180000000013 0 0 -5000 100 1000 2000 11963 0 1000 1000 20968
qm = queryCoordRe.match(mv[0])
if (qm is not None):
g = qm.groups()
qname = g[0]
qfragstart = int(g[1])
qfragend = int(g[2])
else:
qname = mv[0]
qfragstart = 0
qfragend = int(mv[11])
tname = mv[1]
qstrand = int(mv[2])
tstrand = int(mv[3])
pctident = float(mv[5])
tstart = int(mv[6])
tend = int(mv[7])
tlen = int(mv[8])
qstart = int(mv[9])
qend = int(mv[10])
qlen = int(mv[11])
return Match(qname, qfragstart, qfragend, qstrand, qstart, qend, qlen, tname, tstrand, tstart, tend, tlen, pctident)
def SetPlotCoordinates(match, querycs, targetcs, queryi, targeti):
qi = queryi[match.qname]
ti = targeti[match.tname]
queryRangeStart= querycs[qi]
queryRangeEnd = querycs[qi+1]
targetRangeStart = targetcs[ti]
targetRangeEnd = targetcs[ti+1]
# Shift the query coordinates to be relative to the
qStart = match.qfragstart + match.qstart
qEnd = match.qfragstart + match.qend
# now handle strand orientation
tStart = match.tstart
tEnd = match.tend
if (match.tstrand == 1):
tStart = match.tlen - match.tend
tEnd = match.tlen - match.tstart
temp = qStart
qStart = qEnd
qEnd = temp
tStart += targetRangeStart
tEnd += targetRangeStart
qStart += queryRangeStart
qEnd += queryRangeStart
match.plTStart = tStart
match.plTEnd = tEnd
match.plQStart = qStart
match.plQEnd = qEnd
matches = []
for matchLine in mf:
m = ParseMatchLine(matchLine)
if (m is not None):
SetPlotCoordinates(m, queryCumSum, targetCumSum, queryi, targeti)
matches.append(m)
cmap = plt.get_cmap('jet')
#gs = gridspec.GridSpec(len(query), len(target), width_ratios=targetLengths, height_ratios=queryLengths)
gsi = 0
rects = [mlines.Line2D([m.plTStart, m.plTEnd], [m.plQStart, m.plQEnd], color=cmap(1 + math.log(m.pctident/100.0)), lw=2.0) for m in matches]
fig, ax = plt.subplots()
for q in range(len(query)):
for t in range(len(target)):
for i in range(len(matches)):
plt.tight_layout()
if (queryi[m.qname] == q and targeti[m.tname] == t):
ax.add_line(rects[i])
gsi+=1
ax.set_xticks(tLayoutPos)
ax.set_xticklabels(tLayoutText, rotation=90)
ax.set_yticks(qLayoutPos)
ax.set_yticklabels(qLayoutText)
plt.xlim(xmin=0, xmax=targetCumSum[-1]*1.01)
plt.ylim(ymin=0, ymax=queryCumSum[-1]*1.01)
for t in targetCumSum:
plt.axvline(t, linestyle='--', color='black')
for q in queryCumSum:
plt.axhline(q, linestyle='--', color='black')
if (args.savefig is not None):
plt.savefig(args.savefig)
for i in range(len(target)):
plt.figtext(tFrac[i], 0.0, target[i].name)
for tmpFile in tmpFiles:
command = "/bin/rm -f " + tmpFile
try:
subprocess.call(command.split())
except:
print "Cannot delete " + tmpFile
plt.show()