#include "HitWrapper.h"
#include "BamWriter.h"
+#include "WriteResults.h"
+
using namespace std;
const double STOP_CRITERIA = 0.001;
char refName[STRLEN], outName[STRLEN];
char imdName[STRLEN], statName[STRLEN];
-char refF[STRLEN], groupF[STRLEN], cntF[STRLEN], tiF[STRLEN];
+char refF[STRLEN], cntF[STRLEN], tiF[STRLEN];
char mparamsF[STRLEN];
char modelF[STRLEN], thetaF[STRLEN];
double *probv, **countvs;
Refs refs;
-GroupInfo gi;
Transcripts transcripts;
ModelParams mparams;
+bool hasSeed;
+seedType seed;
+
template<class ReadType, class HitType, class ModelType>
void init(ReadReader<ReadType> **&readers, HitContainer<HitType> **&hitvs, double **&ncpvs, ModelType **&mhps) {
READ_INT_TYPE nReads;
return NULL;
}
-template<class ModelType>
-void calcExpectedEffectiveLengths(ModelType& model) {
- int lb, ub, span;
- double *pdf = NULL, *cdf = NULL, *clen = NULL; // clen[i] = sigma_{j=1}^{i}pdf[i]*(lb+i)
-
- model.getGLD().copyTo(pdf, cdf, lb, ub, span);
- clen = new double[span + 1];
- clen[0] = 0.0;
- for (int i = 1; i <= span; i++) {
- clen[i] = clen[i - 1] + pdf[i] * (lb + i);
- }
-
- eel.clear();
- eel.resize(M + 1, 0.0);
- for (int i = 1; i <= M; i++) {
- int totLen = refs.getRef(i).getTotLen();
- int fullLen = refs.getRef(i).getFullLen();
- int pos1 = max(min(totLen - fullLen + 1, ub) - lb, 0);
- int pos2 = max(min(totLen, ub) - lb, 0);
-
- if (pos2 == 0) { eel[i] = 0.0; continue; }
-
- eel[i] = fullLen * cdf[pos1] + ((cdf[pos2] - cdf[pos1]) * (totLen + 1) - (clen[pos2] - clen[pos1]));
- assert(eel[i] >= 0);
- if (eel[i] < MINEEL) { eel[i] = 0.0; }
- }
-
- delete[] pdf;
- delete[] cdf;
- delete[] clen;
-}
-
template<class ModelType>
void writeResults(ModelType& model, double* counts) {
- double denom;
- char outF[STRLEN];
- FILE *fo;
-
- sprintf(modelF, "%s.model", statName);
- model.write(modelF);
-
- //calculate tau values
- double *tau = new double[M + 1];
- memset(tau, 0, sizeof(double) * (M + 1));
-
- denom = 0.0;
- for (int i = 1; i <= M; i++)
- if (eel[i] >= EPSILON) {
- tau[i] = theta[i] / eel[i];
- denom += tau[i];
- }
-
- general_assert(denom > 0, "No alignable reads?!");
-
- for (int i = 1; i <= M; i++) {
- tau[i] /= denom;
- }
-
- //isoform level results
- sprintf(outF, "%s.iso_res", imdName);
- fo = fopen(outF, "w");
- for (int i = 1; i <= M; i++) {
- const Transcript& transcript = transcripts.getTranscriptAt(i);
- fprintf(fo, "%s%c", transcript.getTranscriptID().c_str(), (i < M ? '\t' : '\n'));
- }
- for (int i = 1; i <= M; i++)
- fprintf(fo, "%.2f%c", counts[i], (i < M ? '\t' : '\n'));
- for (int i = 1; i <= M; i++)
- fprintf(fo, "%.15g%c", tau[i], (i < M ? '\t' : '\n'));
- for (int i = 1; i <= M; i++) {
- const Transcript& transcript = transcripts.getTranscriptAt(i);
- fprintf(fo, "%s%c", transcript.getGeneID().c_str(), (i < M ? '\t' : '\n'));
- }
- fclose(fo);
-
- //gene level results
- sprintf(outF, "%s.gene_res", imdName);
- fo = fopen(outF, "w");
- for (int i = 0; i < m; i++) {
- const string& gene_id = transcripts.getTranscriptAt(gi.spAt(i)).getGeneID();
- fprintf(fo, "%s%c", gene_id.c_str(), (i < m - 1 ? '\t' : '\n'));
- }
- for (int i = 0; i < m; i++) {
- double sumC = 0.0; // sum of counts
- int b = gi.spAt(i), e = gi.spAt(i + 1);
- for (int j = b; j < e; j++) sumC += counts[j];
- fprintf(fo, "%.2f%c", sumC, (i < m - 1 ? '\t' : '\n'));
- }
- for (int i = 0; i < m; i++) {
- double sumT = 0.0; // sum of tau values
- int b = gi.spAt(i), e = gi.spAt(i + 1);
- for (int j = b; j < e; j++) sumT += tau[j];
- fprintf(fo, "%.15g%c", sumT, (i < m - 1 ? '\t' : '\n'));
- }
- for (int i = 0; i < m; i++) {
- int b = gi.spAt(i), e = gi.spAt(i + 1);
- for (int j = b; j < e; j++) {
- fprintf(fo, "%s%c", transcripts.getTranscriptAt(j).getTranscriptID().c_str(), (j < e - 1 ? ',' : (i < m - 1 ? '\t' :'\n')));
- }
- }
- fclose(fo);
-
- delete[] tau;
-
- if (verbose) { printf("Expression Results are written!\n"); }
+ sprintf(modelF, "%s.model", statName);
+ model.write(modelF);
+ writeResultsEM(M, refName, imdName, transcripts, theta, eel, countvs[0]);
}
template<class ReadType, class HitType, class ModelType>
double sum;
double bChange = 0.0, change = 0.0; // bChange : biggest change
- READ_INT_TYPE totNum = 0;
+ int totNum = 0;
ModelType model(mparams); //master model
ReadReader<ReadType> **readers;
Params fparams[nThreads];
pthread_t threads[nThreads];
pthread_attr_t attr;
- void *status;
int rc;
}
for (int i = 0; i < nThreads; i++) {
- rc = pthread_join(threads[i], &status);
+ rc = pthread_join(threads[i], NULL);
pthread_assert(rc, "pthread_join", "Cannot join thread " + itos(i) + " (numbered from 0) at ROUND " + itos(ROUND) + "!");
}
if (bChange < change) bChange = change;
}
- if (verbose) { cout<< "ROUND = "<< ROUND<< ", SUM = "<< setprecision(15)<< sum<< ", bChange = " << setprecision(6)<< bChange<< ", totNum = %" << totNum<< endl; }
+ if (verbose) { cout<< "ROUND = "<< ROUND<< ", SUM = "<< setprecision(15)<< sum<< ", bChange = " << setprecision(6)<< bChange<< ", totNum = " << totNum<< endl; }
} while (ROUND < MIN_ROUND || (totNum > 0 && ROUND < MAX_ROUND));
// } while (ROUND < 1);
pthread_assert(rc, "pthread_create", "Cannot create thread " + itos(i) + " (numbered from 0) when generating files for Gibbs sampler!");
}
for (int i = 0; i < nThreads; i++) {
- rc = pthread_join(threads[i], &status);
+ rc = pthread_join(threads[i], NULL);
pthread_assert(rc, "pthread_join", "Cannot join thread " + itos(i) + " (numbered from 0) when generating files for Gibbs sampler!");
}
}
fout.close();
}
- sprintf(thetaF, "%s.theta", statName);
- fo = fopen(thetaF, "w");
- fprintf(fo, "%d\n", M + 1);
-
- // output theta'
- for (int i = 0; i < M; i++) fprintf(fo, "%.15g ", theta[i]);
- fprintf(fo, "%.15g\n", theta[M]);
-
- //calculate expected effective lengths for each isoform
- calcExpectedEffectiveLengths<ModelType>(model);
-
- //correct theta vector
- sum = theta[0];
- for (int i = 1; i <= M; i++)
- if (eel[i] < EPSILON) { theta[i] = 0.0; }
- else sum += theta[i];
-
- general_assert(sum >= EPSILON, "No Expected Effective Length is no less than" + ftos(MINEEL, 6) + "?!");
-
- for (int i = 0; i <= M; i++) theta[i] /= sum;
-
//calculate expected weights and counts using learned parameters
+ //just use the raw theta learned from the data, do not correct for eel or mw
updateModel = false; calcExpectedWeights = true;
for (int i = 0; i <= M; i++) probv[i] = theta[i];
for (int i = 0; i < nThreads; i++) {
pthread_assert(rc, "pthread_create", "Cannot create thread " + itos(i) + " (numbered from 0) when calculating expected weights!");
}
for (int i = 0; i < nThreads; i++) {
- rc = pthread_join(threads[i], &status);
+ rc = pthread_join(threads[i], NULL);
pthread_assert(rc, "pthread_join", "Cannot join thread " + itos(i) + " (numbered from 0) when calculating expected weights!");
}
model.setNeedCalcConPrb(false);
/* destroy attribute */
pthread_attr_destroy(&attr);
- //convert theta' to theta
- double *mw = model.getMW();
- sum = 0.0;
- for (int i = 0; i <= M; i++) {
- theta[i] = (mw[i] < EPSILON ? 0.0 : theta[i] / mw[i]);
- sum += theta[i];
- }
- assert(sum >= EPSILON);
- for (int i = 0; i <= M; i++) theta[i] /= sum;
+
+ sprintf(thetaF, "%s.theta", statName);
+ fo = fopen(thetaF, "w");
+ fprintf(fo, "%d\n", M + 1);
+
+ // output theta'
+ for (int i = 0; i < M; i++) fprintf(fo, "%.15g ", theta[i]);
+ fprintf(fo, "%.15g\n", theta[M]);
+
+ //calculate expected effective lengths for each isoform
+ calcExpectedEffectiveLengths<ModelType>(M, refs, model, eel);
+ polishTheta(M, theta, eel, model.getMW());
// output theta
for (int i = 0; i < M; i++) fprintf(fo, "%.15g ", theta[i]);
READ_INT_TYPE local_N;
HIT_INT_TYPE fr, to, len, id;
vector<double> arr;
- uniform01 rg(engine_type(time(NULL)));
+ engine_type engine(hasSeed ? seed : time(NULL));
+ uniform_01_dist uniform_01;
+ uniform_01_generator rg(engine, uniform_01);
if (verbose) cout<< "Begin to sample reads from their posteriors."<< endl;
for (int i = 0; i < nThreads; i++) {
ifstream fin;
bool quiet = false;
- if (argc < 5) {
- printf("Usage : rsem-run-em refName read_type sampleName sampleToken [-p #Threads] [-b samInpType samInpF has_fn_list_? [fn_list]] [-q] [--gibbs-out] [--sampling]\n\n");
+ if (argc < 6) {
+ printf("Usage : rsem-run-em refName read_type sampleName imdName statName [-p #Threads] [-b samInpType samInpF has_fn_list_? [fn_list]] [-q] [--gibbs-out] [--sampling] [--seed seed]\n\n");
printf(" refName: reference name\n");
printf(" read_type: 0 single read without quality score; 1 single read with quality score; 2 paired-end read without quality score; 3 paired-end read with quality score.\n");
printf(" sampleName: sample's name, including the path\n");
printf(" -q: set it quiet\n");
printf(" --gibbs-out: generate output file used by Gibbs sampler. (default: off)\n");
printf(" --sampling: sample each read from its posterior distribution when bam file is generated. (default: off)\n");
+ printf(" --seed uint32: the seed used for the BAM sampling. (default: off)\n");
printf("// model parameters should be in imdName.mparams.\n");
exit(-1);
}
strcpy(refName, argv[1]);
read_type = atoi(argv[2]);
strcpy(outName, argv[3]);
- sprintf(imdName, "%s.temp/%s", argv[3], argv[4]);
- sprintf(statName, "%s.stat/%s", argv[3], argv[4]);
+ strcpy(imdName, argv[4]);
+ strcpy(statName, argv[5]);
nThreads = 1;
bamSampling = false;
genGibbsOut = false;
pt_fn_list = pt_chr_list = NULL;
+ hasSeed = false;
- for (int i = 5; i < argc; i++) {
+ for (int i = 6; i < argc; i++) {
if (!strcmp(argv[i], "-p")) { nThreads = atoi(argv[i + 1]); }
if (!strcmp(argv[i], "-b")) {
genBamF = true;
if (!strcmp(argv[i], "-q")) { quiet = true; }
if (!strcmp(argv[i], "--gibbs-out")) { genGibbsOut = true; }
if (!strcmp(argv[i], "--sampling")) { bamSampling = true; }
+ if (!strcmp(argv[i], "--seed")) {
+ hasSeed = true;
+ int len = strlen(argv[i + 1]);
+ seed = 0;
+ for (int k = 0; k < len; k++) seed = seed * 10 + (argv[i + 1][k] - '0');
+ }
}
general_assert(nThreads > 0, "Number of threads should be bigger than 0!");
sprintf(refF, "%s.seq", refName);
refs.loadRefs(refF);
M = refs.getM();
- sprintf(groupF, "%s.grp", refName);
- gi.load(groupF);
- m = gi.getm();
sprintf(tiF, "%s.ti", refName);
transcripts.readFrom(tiF);