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
|
#include "ArgumentParser.h"
#include "FileHeader.h"
#include "misc.h"
#include "MyTimer.h"
#include "SimpleSparse.h"
#include "TagAlignments.h"
#include "transposeFiles.h"
#include "VariationalBayes.h"
#include "common.h"
SimpleSparse* readData(const ArgumentParser &args, long trM){//{{{
/*
As parse(filename,maxreads=None) in python
Python difference:
- missing maxreads check
(abort if more than maxreads reads were processed)
*/
long i,j,num,tid;
double prb;
long Ntotal=0,Nmap=0, M=0;
string readId,strand,blank;
ifstream inFile;
MyTimer timer;
TagAlignments *alignments = new TagAlignments();
// Read alignment probabilities {{{
inFile.open(args.args()[0].c_str());
FileHeader fh(&inFile);
ns_fileHeader::AlignmentFileType format;
if((!fh.probHeader(&Nmap,&Ntotal,&M,&format)) || (Nmap ==0)){//{{{
error("Prob file header read failed.\n");
return NULL;
}//}}}
if(format == ns_fileHeader::OLD_FORMAT){
error("Please use new/log format of Prob file.");
return NULL;
}
message("N mapped: %ld\n",Nmap);
messageF("N total: %ld\n",Ntotal);
if(args.verb())message("Reading alignments.\n");
alignments->init(Nmap,0,M);
long mod=10000;
long bad = 0;
timer.start();
for(i = 0; i < Nmap; i++) {
inFile>>readId>>num;
if(!inFile.good())break;
// message("%s %ld\n",(readId).c_str(),num);
for(j = 0; j < num; j++) {
inFile>>tid>>prb;
if(inFile.fail()){
inFile.clear();
// ignore rest of line
j=num;
// this read goes to noise
tid=0;
// 10 means either 10 or exp(10), but should be still be large enough
prb=10;
bad++;
}
switch(format){
case ns_fileHeader::NEW_FORMAT:
alignments->pushAlignment(tid, prb);
break;
case ns_fileHeader::LOG_FORMAT:
alignments->pushAlignmentL(tid, prb);
break;
default:;
}
}
// ignore rest of line
inFile.ignore(10000000,'\n');
alignments->pushRead();
R_INTERUPT;
if(args.verb() && (i % mod == 0) && (i>0)){
message(" %ld ",i);
timer.split();
mod*=10;
}
}
if(bad>0)warning("Main: %ld reads' alignment information were corrupted.\n",bad);
inFile.close();
long Nhits,NreadsReal;
alignments->finalizeRead(&M, &NreadsReal, &Nhits);
// Increase M based on number of transcripts in trInfo file.
if(M<trM)M = trM;
//}}}
if(i<Nmap)message("Read only %ld reads.\n",NreadsReal);
message("All alignments: %ld\n",Nhits);
messageF("Isoforms: %ld\n",M);
Nmap = NreadsReal;
SimpleSparse *beta = new SimpleSparse(Nmap, M, Nhits);
for(i=0;i<=Nmap;i++)beta->rowStart[i]=alignments->getReadsI(i);
for(i=0;i<Nhits;i++){
beta->val[i]=alignments->getProb(i);
beta->col[i]=alignments->getTrId(i);
}
delete alignments;
return beta;
}//}}}
extern "C" int estimateVBExpression(int *argc, char* argv[]) {//{{{
string programDescription =
"Estimates expression given precomputed probabilities of (observed) reads' alignments.\n\
Uses Variational Bayes algorithm to produce parameters for distribution of relative abundances.\n";
// Set options {{{
ArgumentParser args;
args.init(programDescription,"[prob file]",1);
args.addOptionS("o","outPrefix","outFilePrefix",1,"Prefix for the output files.");
args.addOptionS("O","outType","outputType",0,"Output type (theta, RPKM, counts) of the samples sampled from the distribution.","theta");
args.addOptionS("t","trInfoFile","trInfoFileName",0,"File containing transcript information. (Necessary for RPKM samples)");
args.addOptionL("P","procN","procN",0,"Limit the maximum number of threads to be used.",4);
args.addOptionS("m","method","optMethod",0,"Optimization method (steepest, PR, FR, HS).","FR");
args.addOptionL("s","seed","seed",0,"Random initialization seed.");
args.addOptionL("","maxIter","maxIter",0,"Maximum number of iterations.",(long)1e4);
args.addOptionD("","optLimit","limit",0,"Optimisation limit in terms of minimal gradient or change of bound.",1e-5);
args.addOptionL("","samples","samples",0,"Number of samples to be sampled from the distribution.");
args.addOptionB("V","veryVerbose","veryVerbose",0,"More verbose output, better if output forwarded into file.");
if(!args.parse(*argc,argv))return 0;
if(args.verbose)buildTime(argv[0],__DATE__,__TIME__);
OPT_TYPE optM;
if(args.isSet("optMethod")){
if((args.getLowerS("optMethod")=="steepest")||
(args.getLowerS("optMethod")=="vbem"))optM = OPTT_STEEPEST;
else if(args.getLowerS("optMethod")=="pr")optM = OPTT_PR;
else if(args.getLowerS("optMethod")=="fr")optM = OPTT_FR;
else if(args.getLowerS("optMethod")=="hs")optM = OPTT_HS;
else optM = OPTT_FR;
}else optM = OPTT_FR;
args.updateS("outputType", ns_expression::getOutputType(args, "theta"));
if(args.getS("outputType") == "tau"){
error("Main: 'tau' is not valid output type.\n");
return 1;
}
// }}}
MyTimer timer;
timer.start(2);
long M = 0;
SimpleSparse *beta;
TranscriptInfo trInfo;
// {{{ Read transcriptInfo and .prob file
if((!args.isSet("trInfoFileName"))||(!trInfo.readInfo(args.getS("trInfoFileName")))){
if(args.isSet("samples") && (args.getL("samples")>0) && (args.getS("outputType") == "rpkm")){
error("Main: Missing transcript info file. The file is necessary for producing RPKM samples.\n");
return 1;
}
}else{
M = trInfo.getM()+1;
}
beta = readData(args,M);
if(! beta){
error("Main: Reading probabilities failed.\n");
return 1;
}
M = beta->M;
if(M<=0){
error("Main: Invalid number of transcripts in .prob file.\n");
return 1;
}
// }}}
if(args.verbose)timer.split();
if(args.verbose)message("Initializing VB.\n");
VariationalBayes varB(beta,NULL,ns_misc::getSeed(args),args.getL("procN"));
if(args.verbose)timer.split();
if(args.verbose)message("Starting VB optimization.\n");
#ifdef LOG_CONV
varB.setLog(args.getS("outFilePrefix")+".convLog",&timer);
#endif
// Optimize:
if(!args.verbose)varB.beQuiet();
varB.optimize(args.flag("veryVerbose"),optM,args.getL("maxIter"),args.getD("limit"),args.getD("limit"));
if(args.verbose){timer.split(0,'m');}
double *alpha = varB.getAlphas();
double alphaSum = 0 ;
long i;
for(i=0;i<M;i++)alphaSum+=alpha[i];
ofstream outF;
if(! ns_misc::openOutput((args.getS("outFilePrefix")+".m_alphas"), &outF)){
return 1;
}
outF<<"# "<<args.args()[0]<<endl;
outF<<"# M "<<M<<"\n"
"# List includes also 'noise' transcript (first line)\n"
"# <alpha> - parameter of Dirichlet distribution\n"
"# <alpha> <beta> - parameters of the marginal Gamma distribution\n"
"# columns: <mean theta> <alpha> <beta>"<<endl;
outF<<scientific;
outF.precision(9);
for(i=0;i<M;i++){
outF<<alpha[i]/alphaSum<<" "<<alpha[i]<<" "<<alphaSum-alpha[i]<<endl;
}
outF.close();
// free memory
delete beta;
delete[] alpha;
if(args.isSet("samples") && (args.getL("samples")>0)){
string outTypeS = args.getS("outputType");
string samplesFName = args.getS("outFilePrefix")+".VB" + outTypeS;
string samplesTmpName = args.getS("outFilePrefix")+".VB"+outTypeS+"TMP";
timer.start(0);
if(args.verbose)messageF("Generating samples into temporary file %s. ",samplesTmpName.c_str());
if(!ns_misc::openOutput(samplesTmpName, &outF)) return 1;
// Samples are generated without the "noise transcript".
outF<<"# M "<<M-1<<" N "<<args.getL("samples")<<endl;
varB.generateSamples(args.getL("samples"), outTypeS, trInfo.getShiftedLengths(), &outF);
outF.close();
if(args.verbose)timer.split(0);
if(transposeFiles(vector<string>(1, samplesTmpName), samplesFName, args.verbose, "")){
if(args.verbose)message("Removing temporary file %s.\n", samplesTmpName.c_str());
remove(samplesTmpName.c_str());
}else {
error("Main: Transposing samples failed.\n");
return 1;
}
}
if(args.verbose){message("DONE. "); timer.split(2,'m');}
return 0;
}//}}}
#ifndef BIOC_BUILD
int main(int argc, char* argv[]) {
return estimateVBExpression(&argc,argv);
}
#endif
|
