--- /dev/null
+/*
+ * mothurmetastats.cpp
+ * Mothur
+ *
+ * Created by westcott on 7/6/11.
+ * Copyright 2011 Schloss Lab. All rights reserved.
+ *
+ */
+
+#include "mothurmetastats.h"
+#include "mothurfisher.h"
+#include "spline.h"
+
+/***********************************************************/
+MothurMetastats::MothurMetastats(double t, int n) {
+ try {
+ m = MothurOut::getInstance();
+ threshold = t;
+ numPermutations = n;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "MothurMetastats");
+ exit(1);
+ }
+}
+/***********************************************************/
+MothurMetastats::~MothurMetastats() {}
+/***********************************************************/
+ //main metastats function
+int MothurMetastats::runMetastats(string outputFileName, vector< vector<double> >& data, int secGroupingStart) {
+ try {
+ row = data.size(); //numBins
+ column = data[0].size(); //numGroups in subset
+ secondGroupingStart = secGroupingStart; //g
+
+ vector< vector<double> > Pmatrix; Pmatrix.resize(row);
+ for (int i = 0; i < row; i++) { Pmatrix[i].resize(column, 0.0); } // the relative proportion matrix
+ vector< vector<double> > C1; C1.resize(row);
+ for (int i = 0; i < row; i++) { C1[i].resize(3, 0.0); } // statistic profiles for class1 and class 2
+ vector< vector<double> > C2; C2.resize(row); // mean[1], variance[2], standard error[3]
+ for (int i = 0; i < row; i++) { C2[i].resize(3, 0.0); }
+ vector<double> T_statistics; T_statistics.resize(row, 1); // a place to store the true t-statistics
+ vector<double> pvalues; pvalues.resize(row, 1); // place to store pvalues
+ vector<double> qvalues; qvalues.resize(row, 1); // stores qvalues
+
+ //*************************************
+ // convert to proportions
+ // generate Pmatrix
+ //*************************************
+ vector<double> totals; totals.resize(column, 0); // sum of columns
+ //total[i] = total abundance for group[i]
+ for (int i = 0; i < column; i++) {
+ for (int j = 0; j < row; j++) {
+ totals[i] += data[j][i];
+ }
+ }
+
+ for (int i = 0; i < column; i++) {
+ for (int j = 0; j < row; j++) {
+ Pmatrix[j][i] = data[j][i]/totals[i];
+
+ }
+ }
+
+ //#********************************************************************************
+ //# ************************** STATISTICAL TESTING ********************************
+ //#********************************************************************************
+
+ if (column == 2){ //# then we have a two sample comparison
+ //#************************************************************
+ //# generate p values fisher's exact test
+ //#************************************************************
+ double total1, total2;
+ //total for first grouping
+ for (int i = 0; i < secondGroupingStart; i++) { total1 += totals[i]; }
+
+ //total for second grouping
+ for (int i = secondGroupingStart; i < column; i++) { total2 += totals[i]; }
+
+ vector<double> fish; fish.resize(row, 0.0);
+ vector<double> fish2; fish2.resize(row, 0.0);
+
+ for(int i = 0; i < row; i++){
+
+ for(int j = 0; j < secondGroupingStart; j++) { fish[i] += data[i][j]; }
+ for(int j = secondGroupingStart; j < column; j++) { fish2[i] += data[i][j]; }
+
+ double f11, f12, f21, f22;
+ f11 = fish[i];
+ f12 = fish2[i];
+ f21 = total1 - fish[i];
+ f22 = total2 - fish2[i];
+
+ MothurFisher fisher;
+ double pre = fisher.fexact(f11, f12, f21, f22);
+ if (pre > 0.999999999) { pre = 1.0; }
+
+ if (m->control_pressed) { return 1; }
+
+ pvalues[i] = pre;
+ }
+
+ //#*************************************
+ //# calculate q values from p values
+ //#*************************************
+ qvalues = calc_qvalues(pvalues);
+
+ }else { //we have multiple subjects per population
+
+ //#*************************************
+ //# generate statistics mean, var, stderr
+ //#*************************************
+ for(int i = 0; i < row; i++){ // for each taxa
+ //# find the mean of each group
+ double g1Total = 0.0; double g2Total = 0.0;
+ for (int j = 0; j < secondGroupingStart; j++) { g1Total += Pmatrix[i][j]; }
+ C1[i][0] = g1Total/(double)(secondGroupingStart);
+ for (int j = secondGroupingStart; j < column; j++) { g2Total += Pmatrix[i][j]; }
+ C2[i][0] = g2Total/(double)(column-secondGroupingStart);
+
+ //# find the variance of each group
+ double g1Var = 0.0; double g2Var = 0.0;
+ for (int j = 0; j < secondGroupingStart; j++) { g1Var += pow((Pmatrix[i][j]-C1[i][0]), 2); }
+ C1[i][1] = g1Var/(double)(secondGroupingStart-1);
+ for (int j = secondGroupingStart; j < column; j++) { g2Var += pow((Pmatrix[i][j]-C2[i][0]), 2); }
+ C2[i][1] = g2Var/(double)(column-secondGroupingStart-1);
+
+ //# find the std error of each group -std err^2 (will change to std err at end)
+ C1[i][2] = C1[i][1]/(double)(secondGroupingStart);
+ C2[i][2] = C2[i][1]/(double)(column-secondGroupingStart);
+ }
+
+ //#*************************************
+ //# two sample t-statistics
+ //#*************************************
+ for(int i = 0; i < row; i++){ // # for each taxa
+ double xbar_diff = C1[i][0] - C2[i][0];
+ double denom = sqrt(C1[i][2] + C2[i][2]);
+ T_statistics[i] = xbar_diff/denom; // calculate two sample t-statistic
+ }
+
+ /*for (int i = 0; i < row; i++) {
+ for (int j = 0; j < 3; j++) {
+ cout << "C1[" << i+1 << "," << j+1 << "]=" << C1[i][j] << ";" << endl;
+ cout << "C2[" << i+1 << "," << j+1 << "]=" << C2[i][j] << ";" << endl;
+ }
+ cout << "T_statistics[" << i+1 << "]=" << T_statistics[i] << ";" << endl;
+ }*/
+ //#*************************************
+ //# generate initial permuted p-values
+ //#*************************************
+ pvalues = permuted_pvalues(Pmatrix, T_statistics, data);
+
+ //#*************************************
+ //# generate p values for sparse data
+ //# using fisher's exact test
+ //#*************************************
+ double total1, total2;
+ //total for first grouping
+ for (int i = 0; i < secondGroupingStart; i++) { total1 += totals[i]; }
+
+ //total for second grouping
+ for (int i = secondGroupingStart; i < column; i++) { total2 += totals[i]; }
+
+ vector<double> fish; fish.resize(row, 0.0);
+ vector<double> fish2; fish2.resize(row, 0.0);
+
+ for(int i = 0; i < row; i++){
+
+ for(int j = 0; j < secondGroupingStart; j++) { fish[i] += data[i][j]; }
+ for(int j = secondGroupingStart; j < column; j++) { fish2[i] += data[i][j]; }
+
+ if ((fish[1] < secondGroupingStart) && (fish2[i] < (column-secondGroupingStart))) {
+ double f11, f12, f21, f22;
+ f11 = fish[i];
+ f12 = fish2[i];
+ f21 = total1 - fish[i];
+ f22 = total2 - fish2[i];
+
+ MothurFisher fisher;
+ double pre = fisher.fexact(f11, f12, f21, f22);
+ if (pre > 0.999999999) { pre = 1.0; }
+
+ if (m->control_pressed) { return 1; }
+
+ pvalues[i] = pre;
+ }
+ }
+
+ //#*************************************
+ //# calculate q values from p values
+ //#*************************************
+ qvalues = calc_qvalues(pvalues);
+
+ //#*************************************
+ //# convert stderr^2 to std error
+ //#*************************************
+ for(int i = 0; i < row; i++){
+ C1[i][2] = sqrt(C1[i][2]);
+ C2[i][2] = sqrt(C2[i][2]);
+ }
+ }
+
+ // And now we write the files to a text file.
+ struct tm *local;
+ time_t t; t = time(NULL);
+ local = localtime(&t);
+
+ ofstream out;
+ m->openOutputFile(outputFileName, out);
+ out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint);
+
+ out << "Local time and date of test: " << asctime(local) << endl;
+ out << "# rows = " << row << ", # col = " << column << ", g = " << secondGroupingStart << endl << endl;
+ out << numPermutations << " permutations" << endl << endl;
+
+ //output column headings - not really sure... documentation labels 9 columns, there are 10 in the output file
+ //storage 0 = meanGroup1 - line 529, 1 = varGroup1 - line 532, 2 = err rate1 - line 534, 3 = mean of counts group1?? - line 291, 4 = meanGroup2 - line 536, 5 = varGroup2 - line 539, 6 = err rate2 - line 541, 7 = mean of counts group2?? - line 292, 8 = pvalues - line 293
+ out << "OTU\tmean(group1)\tvariance(group1)\tstderr(group1)\tmean(group2)\tvariance(group2)\tstderr(group2)\tp-value\tq-value\n";
+
+ for(int i = 0; i < row; i++){
+ if (m->control_pressed) { out.close(); return 0; }
+
+ //if there are binlabels use them otherwise count.
+ if (m->binLabelsInFile.size() == row) { out << m->binLabelsInFile[i] << '\t'; }
+ else { out << (i+1) << '\t'; }
+
+ out << C1[i][0] << '\t' << C1[i][1] << '\t' << C1[i][2] << '\t' << C2[i][0] << '\t' << C2[i][1] << '\t' << C2[i][2] << '\t' << pvalues[i] << '\t' << qvalues[i] << endl;
+ }
+
+ out << endl << endl;
+ out.close();
+
+
+
+ return 0;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "runMetastats");
+ exit(1);
+ }
+}
+/***********************************************************/
+vector<double> MothurMetastats::permuted_pvalues(vector< vector<double> >& Imatrix, vector<double>& tstats, vector< vector<double> >& Fmatrix) {
+ try {
+ //# matrix stores tstats for each taxa(row) for each permuted trial(column)
+ vector<double> ps; ps.resize(row, 0.0); //# to store the pvalues
+ vector< vector<double> > permuted_ttests; permuted_ttests.resize(numPermutations);
+ for (int i = 0; i < numPermutations; i++) { permuted_ttests[i].resize(row, 0.0); }
+
+ //# calculate null version of tstats using B permutations.
+ for (int i = 0; i < numPermutations; i++) {
+ permuted_ttests[i] = permute_and_calc_ts(Imatrix);
+ }
+
+ //# calculate each pvalue using the null ts
+ if ((secondGroupingStart) < 8 || (column-secondGroupingStart) < 8){
+ vector< vector<double> > cleanedpermuted_ttests; cleanedpermuted_ttests.resize(numPermutations); //# the array pooling just the frequently observed ts
+ //# then pool the t's together!
+ //# count how many high freq taxa there are
+ int hfc = 1;
+ for (int i = 0; i < row; i++) { // # for each taxa
+ double group1Total = 0.0; double group2Total = 0.0;
+ for(int j = 0; j < secondGroupingStart; j++) { group1Total += Fmatrix[i][j]; }
+ for(int j = secondGroupingStart; j < column; j++) { group2Total += Fmatrix[i][j]; }
+
+ if (group1Total >= secondGroupingStart || group2Total >= (column-secondGroupingStart)){
+ hfc++;
+ for (int j = 0; j < numPermutations; j++) { cleanedpermuted_ttests[j].push_back(permuted_ttests[j][i]); }
+ }
+ }
+
+ //#now for each taxa
+ for (int i = 0; i < row; i++) {
+ //number of cleanedpermuted_ttests greater than tstat[i]
+ int numGreater = 0;
+ for (int j = 0; j < numPermutations; j++) {
+ for (int k = 0; k < hfc; k++) {
+ if (cleanedpermuted_ttests[j][k] > abs(tstats[i])) { numGreater++; }
+ }
+ }
+
+ ps[i] = (1/(double)(numPermutations*hfc))*numGreater;
+ }
+ }else{
+ for (int i = 0; i < row; i++) {
+ //number of permuted_ttests[i] greater than tstat[i] //(sum(permuted_ttests[i,] > abs(tstats[i]))+1)
+ int numGreater = 1;
+ for (int j = 0; j < numPermutations; j++) { if (permuted_ttests[j][i] > abs(tstats[i])) { numGreater++; } }
+ ps[i] = (1/(double)(numPermutations+1))*numGreater;
+ }
+ }
+
+ return ps;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "permuted_pvalues");
+ exit(1);
+ }
+}
+/***********************************************************/
+vector<double> MothurMetastats::permute_and_calc_ts(vector< vector<double> >& Imatrix) {
+ try {
+ vector< vector<double> > permutedMatrix = Imatrix;
+
+ //randomize columns, ie group abundances.
+ for (int i = 0; i < permutedMatrix.size(); i++) { random_shuffle(permutedMatrix[i].begin(), permutedMatrix[i].end()); }
+
+ //calc ts
+ vector< vector<double> > C1; C1.resize(row);
+ for (int i = 0; i < row; i++) { C1[i].resize(3, 0.0); } // statistic profiles for class1 and class 2
+ vector< vector<double> > C2; C2.resize(row); // mean[1], variance[2], standard error[3]
+ for (int i = 0; i < row; i++) { C2[i].resize(3, 0.0); }
+ vector<double> Ts; Ts.resize(row, 0.0); // a place to store the true t-statistics
+
+ //#*************************************
+ //# generate statistics mean, var, stderr
+ //#*************************************
+ for(int i = 0; i < row; i++){ // for each taxa
+ //# find the mean of each group
+ double g1Total = 0.0; double g2Total = 0.0;
+ for (int j = 0; j < secondGroupingStart; j++) { g1Total += permutedMatrix[i][j]; }
+ C1[i][0] = g1Total/(double)(secondGroupingStart);
+ for (int j = secondGroupingStart; j < column; j++) { g2Total += permutedMatrix[i][j]; }
+ C2[i][0] = g2Total/(double)(column-secondGroupingStart);
+
+ //# find the variance of each group
+ double g1Var = 0.0; double g2Var = 0.0;
+ for (int j = 0; j < secondGroupingStart; j++) { g1Var += pow((permutedMatrix[i][j]-C1[i][0]), 2); }
+ C1[i][1] = g1Var/(double)(secondGroupingStart-1);
+ for (int j = secondGroupingStart; j < column; j++) { g2Var += pow((permutedMatrix[i][j]-C2[i][0]), 2); }
+ C2[i][1] = g2Var/(double)(column-secondGroupingStart-1);
+
+ //# find the std error of each group -std err^2 (will change to std err at end)
+ C1[i][2] = C1[i][1]/(double)(secondGroupingStart);
+ C2[i][2] = C2[i][1]/(double)(column-secondGroupingStart);
+ }
+
+ //#*************************************
+ //# two sample t-statistics
+ //#*************************************
+ for(int i = 0; i < row; i++){ // # for each taxa
+ double xbar_diff = C1[i][0] - C2[i][0];
+ double denom = sqrt(C1[i][2] + C2[i][2]);
+ Ts[i] = abs(xbar_diff/denom); // calculate two sample t-statistic
+ }
+
+ return Ts;
+
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "permuted_ttests");
+ exit(1);
+ }
+}
+/***********************************************************/
+vector<double> MothurMetastats::calc_qvalues(vector<double>& pValues) {
+ try {
+
+ /* cout << "x <- c(" << pValues[0];
+ for (int l = 1; l < pValues.size(); l++){
+ cout << ", " << pValues[l];
+ }
+ cout << ")\n";*/
+
+ int numRows = pValues.size();
+ vector<double> qvalues(numRows, 0.0);
+
+ //fill lambdas with 0.00, 0.01, 0.02... 0.95
+ vector<double> lambdas(96, 0);
+ for (int i = 1; i < lambdas.size(); i++) { lambdas[i] = lambdas[i-1] + 0.01; }
+
+ vector<double> pi0_hat(lambdas.size(), 0);
+
+ //calculate pi0_hat
+ for (int l = 0; l < lambdas.size(); l++){ // for each lambda value
+ int count = 0;
+ for (int i = 0; i < numRows; i++){ // for each p-value in order
+ if (pValues[i] > lambdas[l]){ count++; }
+ }
+ pi0_hat[l] = count/(double)(numRows*(1-lambdas[l]));
+ }
+
+ double pi0 = smoothSpline(lambdas, pi0_hat, 3);
+
+ //order p-values
+ vector<double> ordered_qs = qvalues;
+ vector<int> ordered_ps(pValues.size(), 0);
+ for (int i = 1; i < ordered_ps.size(); i++) { ordered_ps[i] = ordered_ps[i-1] + 1; }
+ vector<double> tempPvalues = pValues;
+ OrderPValues(0, numRows-1, tempPvalues, ordered_ps);
+
+ ordered_qs[numRows-1] = min((pValues[ordered_ps[numRows-1]]*pi0), 1.0);
+ for (int i = (numRows-2); i >= 0; i--){
+ double p = pValues[ordered_ps[i]];
+ double temp = p*numRows*pi0/(double)(i+1);
+
+ ordered_qs[i] = min(temp, ordered_qs[i+1]);
+ }
+
+ //re-distribute calculated qvalues to appropriate rows
+ for (int i = 0; i < numRows; i++){
+ qvalues[ordered_ps[i]] = ordered_qs[i];
+ }
+
+ return qvalues;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "calc_qvalues");
+ exit(1);
+ }
+}
+/***********************************************************/
+int MothurMetastats::OrderPValues(int low, int high, vector<double>& p, vector<int>& order) {
+ try {
+
+ if (low < high) {
+ int i = low+1;
+ int j = high;
+ int pivot = (low+high) / 2;
+
+ swapElements(low, pivot, p, order); //puts pivot in final spot
+
+ /* compare value */
+ double key = p[low];
+
+ /* partition */
+ while(i <= j) {
+ /* find member above ... */
+ while((i <= high) && (p[i] <= key)) { i++; }
+
+ /* find element below ... */
+ while((j >= low) && (p[j] > key)) { j--; }
+
+ if(i < j) {
+ swapElements(i, j, p, order);
+ }
+ }
+
+ swapElements(low, j, p, order);
+
+ /* recurse */
+ OrderPValues(low, j-1, p, order);
+ OrderPValues(j+1, high, p, order);
+ }
+
+ return 0;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "OrderPValues");
+ exit(1);
+ }
+}
+/***********************************************************/
+int MothurMetastats::swapElements(int i, int j, vector<double>& p, vector<int>& order) {
+ try {
+
+ double z = p[i];
+ p[i] = p[j];
+ p[j] = z;
+
+ int temp = order[i];
+ order[i] = order[j];
+ order[j] = temp;
+
+ return 0;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "swapElements");
+ exit(1);
+ }
+}
+/***********************************************************/
+double MothurMetastats::smoothSpline(vector<double>& x, vector<double>& y, int df) {
+ try {
+
+ double result = 0.0;
+ int n = x.size();
+ vector<double> w(n, 1);
+ double* xb = new double[n];
+ double* yb = new double[n];
+ double* wb = new double[n];
+ double yssw = 0.0;
+ for (int i = 0; i < n; i++) {
+ wb[i] = w[i];
+ yb[i] = w[i]*y[i];
+ yssw += (w[i] * y[i] * y[i]) - wb[i] * (yb[i] * yb[i]);
+ xb[i] = (x[i] - x[0]) / (x[n-1] - x[0]);
+ }
+
+ vector<double> knot = sknot1(xb, n);
+ int nk = knot.size() - 4;
+
+ double low = -1.5; double high = 1.5; double tol = 1e-04; double eps = 2e-08; int maxit = 500;
+ int ispar = 0; int icrit = 3; double dofoff = 3.0;
+ double penalty = 1.0;
+ int ld4 = 4; int isetup = 0; int ldnk = 1; int ier; double spar = 1.0; double crit;
+
+ double* knotb = new double[knot.size()];
+ double* coef1 = new double[nk];
+ double* lev1 = new double[n];
+ double* sz1 = new double[n];
+ for (int i = 0; i < knot.size(); i++) { knotb[i] = knot[i]; }
+
+ Spline spline;
+ spline.sbart(&penalty, &dofoff, &xb[0], &yb[0], &wb[0], &yssw, &n, &knotb[0], &nk, &coef1[0], &sz1[0], &lev1[0], &crit,
+ &icrit, &spar, &ispar, &maxit, &low, &high, &tol, &eps, &isetup, &ld4, &ldnk, &ier);
+
+ result = coef1[nk-1];
+
+ //free memory
+ delete [] xb;
+ delete [] yb;
+ delete [] wb;
+ delete [] knotb;
+ delete [] coef1;
+ delete [] lev1;
+ delete [] sz1;
+
+ return result;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "smoothSpline");
+ exit(1);
+ }
+}
+/***********************************************************/
+vector<double> MothurMetastats::sknot1(double* x, int n) {
+ try {
+ vector<double> knots;
+ int nk = nkn(n);
+
+ //R equivalent - rep(x[1L], 3L)
+ knots.push_back(x[0]); knots.push_back(x[0]); knots.push_back(x[0]);
+
+ //generate a sequence of nk equally spaced values from 1 to n. R equivalent = seq.int(1, n, length.out = nk)
+ vector<int> indexes = getSequence(0, n-1, nk);
+ for (int i = 0; i < indexes.size(); i++) { knots.push_back(x[indexes[i]]); }
+
+ //R equivalent - rep(x[n], 3L)
+ knots.push_back(x[n-1]); knots.push_back(x[n-1]); knots.push_back(x[n-1]);
+
+ return knots;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "sknot1");
+ exit(1);
+ }
+}
+/***********************************************************/
+vector<int> MothurMetastats::getSequence(int start, int end, int length) {
+ try {
+ vector<int> sequence;
+ double increment = (end-start) / (double) (length-1);
+
+ sequence.push_back(start);
+ for (int i = 1; i < length-1; i++) {
+ sequence.push_back(int(i*increment));
+ }
+ sequence.push_back(end);
+
+ return sequence;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "getSequence");
+ exit(1);
+ }
+}
+/***********************************************************/
+//not right, havent fully figured out the variable types yet...
+int MothurMetastats::nkn(int n) {
+ try {
+
+ if (n < 50) { return n; }
+ else {
+ double a1 = log2(50);
+ double a2 = log2(100);
+ double a3 = log2(140);
+ double a4 = log2(200);
+
+ if (n < 200) {
+ return (int)pow(2.0, (a1 + (a2 - a1) * (n - 50)/(float)150));
+ }else if (n < 800) {
+ return (int)pow(2.0, (a2 + (a3 - a2) * (n - 200)/(float)600));
+ }else if (n < 3200) {
+ return (int)pow(2.0, (a3 + (a4 - a3) * (n - 800)/(float)2400));
+ }else {
+ return (int)pow((double)(200 + (n - 3200)), 0.2);
+ }
+ }
+
+ return 0;
+
+ }catch(exception& e) {
+ m->errorOut(e, "MothurMetastats", "nkn");
+ exit(1);
+ }
+}
+/***********************************************************/
+
+
+
+
+
projects / mothur.git / blobdiff