--- /dev/null
+/* Last changed Time-stamp: <2008-07-04 15:57:03 ulim> */
+/*
+ partiton function for RNA secondary structures
+
+ Ivo L Hofacker
+
+ Vienna RNA package
+*/
+/*
+ $Log: part_func_up.c,v $
+ Revision 1.4 2008/07/04 14:27:36 ivo
+ Modify output (again)
+
+ Revision 1.3 2008/05/08 14:11:55 ivo
+ minor output changes
+
+ Revision 1.2 2007/12/13 10:19:54 ivo
+ major RNAup update from Ulli
+
+ Revision 1.1 2007/04/30 15:13:13 ivo
+ merge RNAup into package
+
+ Revision 1.11 2006/07/17 11:11:43 ulim
+ removed all globals from fold_vars.h,c, cleaned code
+
+ Revision 1.10 2006/07/12 09:19:29 ulim
+ global variables w, incr3 and incr5 are now local
+
+ Revision 1.9 2006/07/11 12:45:02 ulim
+ remove redundancy in function pf_interact(...)
+
+ Revision 1.8 2006/03/08 15:26:37 ulim
+ modified -o[1|2], added meaningful default
+
+ Revision 1.5 2006/01/23 11:27:04 ulim
+ include file into new package version. cleaned it
+
+ Revision 1.2 2005/07/29 15:13:37 ulim
+ put the function, calculating the probability of an unpaired region in
+ an RNA and the function calculating the prob. of interaction between 2 RNAs
+ in a seperate file (pf_two.c)
+
+ Revision 1.1 2005/07/26 13:27:12 ulim
+ Initial revision
+
+ Revision 1.2 2005/07/01 13:14:57 ulim
+ fixed error in scaling, included new commandline options -incr5, -incr3 to
+ allow a variable number of unpaired positions 5' and 3' of the site of
+ interaction between the two RNAs
+
+ Revision 1.1 2005/04/19 08:16:38 ulim
+ Initial revision
+*/
+
+#include <config.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <math.h>
+#include <float.h> /* #defines FLT_MAX ... */
+#include <unistd.h>
+#include "fold.h"
+#include "utils.h"
+#include "energy_par.h"
+#include "fold_vars.h"
+#include "pair_mat.h"
+#include "params.h"
+#include "part_func.h"
+#include "loop_energies.h"
+#include "part_func_up.h"
+#include "duplex.h"
+/*@unused@*/
+static char rcsid[] UNUSED = "$Id: part_func_up.c,v 1.4 2008/07/04 14:27:36 ivo Exp $";
+
+#define CO_TURN 0
+#define ZERO(A) (fabs(A) < DBL_EPSILON)
+#define EQUAL(A,B) (fabs((A)-(B)) < 1000*DBL_EPSILON)
+#define ISOLATED 256.0
+/* #define NUMERIC 1 */
+
+/*
+#################################
+# GLOBAL VARIABLES #
+#################################
+*/
+
+/*
+#################################
+# PRIVATE VARIABLES #
+#################################
+*/
+PRIVATE short *S=NULL, *S1=NULL, *SS=NULL, *SS2=NULL;
+PRIVATE pf_paramT *Pf = NULL;/* use this structure for all the exp-arrays*/
+PRIVATE FLT_OR_DBL *qb=NULL, *qm=NULL, *prpr=NULL; /* add arrays for pf_unpaired()*/
+PRIVATE FLT_OR_DBL *probs=NULL;
+PRIVATE FLT_OR_DBL *q1k=NULL, *qln=NULL;
+PRIVATE double *qqm2=NULL, *qq_1m2=NULL, *qqm=NULL, *qqm1=NULL;
+PRIVATE FLT_OR_DBL *scale=NULL, *expMLbase=NULL;
+PRIVATE char *ptype=NULL; /* precomputed array of pair types */
+PRIVATE int init_length; /* length in last call to init_pf_fold()*/
+PRIVATE double init_temp; /* temperature in last call to scale_pf_params */
+PRIVATE int *my_iindx = NULL;
+/* make iptypes array for intermolecular constrains (ipidx for indexing)*/
+
+
+/*
+#################################
+# PRIVATE FUNCTION DECLARATIONS #
+#################################
+*/
+PRIVATE pu_out *get_u_vals(pu_contrib *p_c,
+ int **unpaired_values,
+ char *select_contrib);
+
+PRIVATE int plot_free_pu_out( pu_out* res,
+ interact *pint,
+ char *ofile,
+ char *head);
+
+PRIVATE void scale_stru_pf_params(unsigned int length);
+
+PRIVATE void init_pf_two(int length);
+
+PRIVATE void scale_int(const char *s,
+ const char *sl,
+ double *sc_int);
+
+PRIVATE void encode_seq( const char *s1,
+ const char *s2);
+
+PRIVATE constrain *get_ptypes(char *S,
+ const char *structure);
+
+PRIVATE void get_up_arrays(unsigned int length);
+
+PRIVATE void free_up_arrays(void);
+
+PRIVATE void set_encoded_seq(const char *sequence,
+ short **S,
+ short **S1);
+
+PRIVATE void get_interact_arrays(unsigned int n1,
+ unsigned int n2,
+ pu_contrib *p_c,
+ pu_contrib *p_c2,
+ int w,
+ int incr5,
+ int incr3,
+ double ***p_c_S,
+ double ***p_c2_S);
+
+/*
+#################################
+# BEGIN OF FUNCTION DEFINITIONS #
+#################################
+*/
+
+PUBLIC pu_contrib *get_pu_contrib_struct(unsigned int n, unsigned int w){
+ unsigned int i;
+ pu_contrib *pu = (pu_contrib *)space(sizeof(pu_contrib));
+ pu->length = n;
+ pu->w = w;
+ /* contributions to probability of being unpaired witihin a(n)
+ H hairpin,
+ I interior loop,
+ M muliloop,
+ E exterior loop*/
+ /* pu_test->X[i][j] where i <= j and i [1...n], j = [1...w[ */
+ pu->H = (double **)space(sizeof(double *) * (n + 1));
+ pu->I = (double **)space(sizeof(double *) * (n + 1));
+ pu->M = (double **)space(sizeof(double *) * (n + 1));
+ pu->E = (double **)space(sizeof(double *) * (n + 1));
+ for(i=0;i<=n;i++){
+ pu->H[i] = (double *)space(sizeof(double) * (w + 1));
+ pu->I[i] = (double *)space(sizeof(double) * (w + 1));
+ pu->M[i] = (double *)space(sizeof(double) * (w + 1));
+ pu->E[i] = (double *)space(sizeof(double) * (w + 1));
+ }
+ return pu;
+}
+
+PUBLIC void free_pu_contrib_struct(pu_contrib *pu){
+ unsigned int i;
+ if(pu != NULL){
+ for(i=0;i<=pu->length;i++){
+ free(pu->H[i]);
+ free(pu->I[i]);
+ free(pu->M[i]);
+ free(pu->E[i]);
+ }
+ free(pu->H);
+ free(pu->I);
+ free(pu->M);
+ free(pu->E);
+ free(pu);
+ }
+}
+
+/* you have to call pf_fold(sequence, structure); befor pf_unstru */
+PUBLIC pu_contrib *pf_unstru(char *sequence, int w){
+ int n, i, j, v, k, l, o, p, ij, kl, po, u, u1, d, type, type_2, tt, uu;
+ unsigned int size;
+ double temp, tqm2;
+ double qbt1, *tmp, Zuij, sum_l, *sum_M;
+ double *store_H, *store_Io, **store_I2o; /* hairp., interior contribs */
+ double *store_M_qm_o,*store_M_mlbase; /* multiloop contributions */
+ pu_contrib *pu_test;
+
+ sum_l = 0.0;
+ temp = 0;
+ n = (int) strlen(sequence);
+ sum_M = (double *) space((n+1) * sizeof(double));
+ pu_test = get_pu_contrib_struct((unsigned)n, (unsigned)w);
+ size = ((n+1)*(n+2))>>1;
+
+ get_up_arrays((unsigned) n);
+ init_pf_two(n);
+
+ /* init everything */
+ for (d=0; d<=TURN; d++)
+ for (i=1; i<=n-d; i++){
+ j=i+d;
+ ij = my_iindx[i]-j;
+ if(d < w) {
+ pu_test->H[i][d]=pu_test->I[i][d]=pu_test->M[i][d]=pu_test->E[i][d]=0.;
+ }
+ }
+
+
+ for (i=0; i<size; i++)
+ prpr[i]= probs[i];
+
+ sum_M[0] = 0.;
+ for (i=1; i<=n; i++){
+ /* set auxillary arrays to 0, reuse qqm and qqm1, reuse qqm2 and qq_1m2*/
+ sum_M[i] = qqm[i] = qqm1[i] = qqm2[i] = qq_1m2[i] = 0;
+ for (j=i+TURN+1; j<=n; j++){
+ ij = my_iindx[i]-j;
+ /* i need the part_func of all structures outside bp[ij] */
+ if(qb[ij] > 0.0) prpr[ij]= (probs[ij]/qb[ij]);
+ }
+ }
+
+ /* alloc even more memory */
+ store_I2o = (double **)space(sizeof(double *) * (n + 1)); /* for p,k */
+ for(i=0;i<=n;i++)
+ store_I2o[i] = (double *)space(sizeof(double) * (MAXLOOP + 2));
+
+ /* expMLbase[i-p]*dangles_po */
+ store_M_mlbase = (double *)space(sizeof(double) * (size + 1));
+
+ /* 2. exterior bp (p,o) encloses unpaired region [i,i+w[*/
+ for (o=TURN+2;o<=n; o++) {
+ double sum_h;
+ /*allocate space for arrays to store different contributions to H, I & M */
+ store_H = (double *)space(sizeof(double) * (o+2));
+ /* unpaired between ]l,o[ */
+ store_Io = (double *)space(sizeof(double) * (o+2));
+ /* qm[p+1,i-1]*dangles_po */
+ store_M_qm_o = (double *)space(sizeof(double) * (n+1));
+
+ for (p=o-TURN-1; p>=1; p--) {
+ /* construction of partition function of segment [p,o], given that
+ an unpaired region [i,i+w[ exists within [p,o] */
+ u = o-p-1;
+ po = my_iindx[p]-o;
+ type = ptype[po];
+ if(type){
+
+ /*hairpin contribution*/
+ if (((type==3)||(type==4))&&no_closingGU)
+ temp = 0.;
+ else
+ temp = prpr[po] * exp_E_Hairpin(u, type, S1[p+1], S1[o-1], sequence+p-1, Pf) * scale[u+2];
+ /* all H contribs are collect for the longest unpaired region */
+ store_H[p+1] = temp;
+
+ /* interior loops with interior pair k,l and an unpaired region of
+ length w between p and k || l and o*/
+ for (k=p+1; k<=MIN2(p+MAXLOOP+1,o-TURN-2); k++) {
+ u1 = k-p-1;
+ sum_l = 0.;
+ for (l=MAX2(k+TURN+1,o-1-MAXLOOP+u1); l<o; l++) {
+ kl = my_iindx[k]-l;
+ type_2 = ptype[kl];
+ if((l+1) < o) store_Io[l+1] += sum_l;
+
+ temp=0.;
+ if (type_2){
+ type_2 = rtype[type_2];
+ temp = prpr[po] * qb[kl] * exp_E_IntLoop(u1, o-l-1, type, type_2, S1[p+1], S1[o-1], S1[k-1], S1[l+1], Pf) *scale[u1+o-l+1];
+ if((l+1) < o) store_Io[l+1] += temp; /* unpaired region between ]l,o[ */
+ sum_l += temp;
+ } /* end of if pair(k,l) */
+ } /* end of l */
+ /* unpaired in region ]p,k[ */
+ for(i=p+1;i <= k-1;i++)
+ store_I2o[i][MIN2(w-1,k-i-1)] += sum_l;
+ } /* end of k */
+ } /*end of if(type) test for bp (p,o) */
+
+ /* multiple stem loop contribution
+ calculate qm2[my_iindx[i]-j] in the course of the calculation
+ of the multiple stem loop contribution:
+ advantage: you save memory:
+ instead of a (n+1)*n array for qqm2 you only need 2*n arrays
+ disadvantage: you have to use two times the op-loop for the full
+ multiloop contribution
+ first op-loop: index o goes from 1...n and
+ index p from o-TURN-1 ... 1
+ second op-loop: index o goes from n...1 and
+ index p from o+TURN+1 ... n !!
+ HERE index o goes from 1...n and index p o-TURN-1 ... 1 ,
+ we calculate the contributions to multiple stem loop
+ where exp(i+w-1-p)*(qqm2 values between i+w and o-1)
+ AND qm[iindex[p+1]-(i-1)]*exp(beta*w)*qm[iindex[i+w]-(o-1)]
+ you have to recalculate of qqm matrix containing final stem
+ contributions to multiple loop partition function
+ from segment p,o */
+
+ /* recalculate qqm[]
+ qqm[p] := (contribution with exact one loop in region (p,o)*/
+ qqm[p] = qqm1[p] * expMLbase[1];
+ if(type){
+ qbt1 = qb[po] * exp_E_MLstem(type, (p>1) ? S1[p-1] : -1, (o<n) ? S1[o+1] : -1, Pf);
+ qqm[p] += qbt1;
+ /* reverse dangles for prpr[po]*... */
+ temp = 0.;
+ tt = rtype[type];
+ temp = prpr[po] * exp_E_MLstem(tt, S1[o-1], S1[p+1], Pf) * scale[2] * Pf->expMLclosing;
+ for(i=p+1; i < o; i++) {
+ int p1i = (p+1) < (i-1) ? my_iindx[p+1]-(i-1) : 0;
+ /*unpaired region expMLbase[i-p] left of structured
+ region qq_1m2[i+1]*/
+ /* @expMLbase: note distance of i-p == i-(p+1)+1 */
+ store_M_mlbase[my_iindx[p+1]-i] += expMLbase[i-p] * temp * qq_1m2[i+1];
+ /* structured region qm[p1i] left of unpaired region */
+ /* contribition for unpaired region is added after the p-loop */
+ store_M_qm_o[i] += qm[p1i] * temp;
+ } /*end of for i ... */
+ }
+
+ for(tqm2 = 0., i=p+1; i < o; i++)
+ tqm2 += qm[my_iindx[p]-i] * qqm[i+1];
+
+ /* qqm2[p] contrib with at least 2 loops in region (p,o) */
+ qqm2[p] = tqm2;
+ } /* end for (p=..) */
+
+ for(sum_h = 0., i=1; i < o; i++) {
+ int max_v, vo;
+ sum_h += store_H[i];
+ max_v = MIN2(w-1,o-i-1);
+ for(v=max_v; v >= 0; v--){
+ /* Hairpins */
+ pu_test->H[i][v] += sum_h;/* store_H[i][v] + store_H[i][max_v]; */
+ /* Interior loops: unpaired region between ]l,o[ calculated here !*/
+ /* unpaired region between ]p,k[ collected after after o-loop */
+ if(v <= MIN2(max_v,MAXLOOP)) {
+ pu_test->I[i][v] += store_Io[i]; /* ]l,o[ */
+ }
+ /* Multiloops:*/
+ /* unpaired region [i,v] between structured regions ]p,i[ and ]v,o[. */
+ /* store_M_qm_o[i] = part. funct over all structured regions ]p,i[ */
+ vo = (i+v+1) <= (o-1) ? my_iindx[i+v+1]-(o-1): 0;
+ pu_test->M[i][v] += store_M_qm_o[i]*expMLbase[v+1]*qm[vo];
+ }
+ }
+ tmp = qqm1; qqm1=qqm; qqm=tmp;
+ tmp = qqm2; qqm2=qq_1m2; qq_1m2=tmp;
+
+ free(store_Io);
+ free(store_H);
+ free(store_M_qm_o);
+ }/* end for (o=..) */
+
+ for(i=1; i < n; i++) {
+ int max_v;
+ double sum_iv;
+ sum_iv = 0.;
+ max_v = MIN2(w-1,n-i);
+ for(v=n; v >=0; v--) {
+ if(v <= MIN2(max_v,MAXLOOP)) {
+ /* all unpaired regions [i,v] between p and k in interior loops */
+ /* notice v runs from max_v -> 0, sum_iv sums all int. l. contribs */
+ /* for each x, v < x =< max_v, since they contribute to [i,v] */
+ sum_iv += store_I2o[i][v];
+ pu_test->I[i][v] += sum_iv;
+ }
+ /* all unpaired region [i,v] for a fixed v, given that */
+ /* region ]v,o[ contains at least 2 structures qq_1m2[v+1]; */
+ if(v >= i) {
+ sum_M[v] += store_M_mlbase[my_iindx[i]-v];
+ if(v-i<=max_v) {
+ pu_test->M[i][v-i] += sum_M[v];
+ }
+ }
+ }
+ }
+
+ for(i=0;i<=n;i++) {
+ free(store_I2o[i]);
+ }
+ free(store_I2o);
+
+ for (i=1; i<=n; i++) {
+ /* set auxillary arrays to 0 */
+ qqm[i] = qqm1[i] = qqm2[i] = qq_1m2[i] = 0;
+ }
+
+ /* 2. exterior bp (p,o) encloses unpaired region [i,j]
+ HERE index o goes from n...1 and index p from o+TURN+1 ... n,
+ that is, we add the one multiloop contribution that we
+ could not calculate before */
+
+/* is free'ing plus allocating faster than looping over all entries an setting them to 0? */
+#if 0
+ free(store_M_mlbase);
+ store_M_mlbase = (double *) space(sizeof(double) * (size + 1));
+#else
+ /* this should be the fastest way to set everything to 0 */
+ memset(store_M_mlbase, 0, sizeof(double) * (size + 1));
+#endif
+
+ for (o=n-TURN-1;o>=1; o--) {
+ for (p=o+TURN+1; p<=n; p++) {
+ po = my_iindx[o]-p;
+ type = ptype[po];
+ /* recalculate of qqm matrix containing final stem
+ contributions to multiple loop partition function
+ from segment [o,p] */
+ qqm[p] = qqm1[p] * expMLbase[1];
+ if (type) {
+ qbt1 = qb[po];
+ qbt1 *= exp_E_MLstem(type, (o>1) ? S1[o-1] : -1, (p<n) ? S1[p+1] : -1, Pf);
+ qqm[p] += qbt1;
+ /* revers dangles for prpr[po]... */
+ temp=0.;
+ tt=rtype[type];
+ temp = prpr[po]*exp_E_MLstem(tt, S1[p-1], S1[o+1], Pf) * Pf->expMLclosing * scale[2];
+ }
+ tqm2=0.;
+ for(i=o+1; i < p; i++) {
+ uu=0;
+ tqm2+=qqm[i]*qm[my_iindx[i+1]-p];
+
+ if(type !=0) {
+ double temp2;
+ temp2=0;
+ /* structured region qq_1m2[i-1] left of unpaired r. expMLbase[p-i]*/
+ /* @expMLbase: note distance of p-i == p+1-i+1 */
+ store_M_mlbase[my_iindx[i]-p+1] += qq_1m2[i-1]*expMLbase[p-i]*temp;
+ }
+ }/*end of for i ....*/
+ qqm2[p] = tqm2;
+ }/* end for (p=..) */
+ tmp = qqm1; qqm1=qqm; qqm=tmp;
+ tmp = qqm2; qqm2=qq_1m2; qq_1m2=tmp;
+ }/* end for (o=..) */
+ /* now collect the missing multiloop contributions */
+ for(i=0;i<=n;i++) { sum_M[i]=0.; }
+ for(i=1; i<=n;i++) {
+ int v_max = MIN2(w-1,n-i);
+ for(v=n; v>=i; v--){
+ sum_M[i] += store_M_mlbase[my_iindx[i]-v];
+ if ((v-i <= v_max) ) {
+ pu_test->M[i][v-i] += sum_M[i];
+ }
+ }
+ }
+
+ /* 1. region [i,j] exterior to all loops */
+ Zuij=0.;
+ for (i=1; i<=n; i++) {
+ uu=0;
+ for(j=i; j<MIN2(i+w,n+1);j++){
+ ij=my_iindx[i]-j;
+ temp=q1k[i-1]*1*scale[j-i+1]*qln[j+1]/q1k[n];
+ pu_test->E[i][j-i]+=temp;
+
+ }
+ }
+
+ free(sum_M);
+ free(store_M_mlbase);
+ free_up_arrays();
+ return pu_test;
+}
+
+
+PRIVATE void get_interact_arrays(unsigned int n1,
+ unsigned int n2,
+ pu_contrib *p_c,
+ pu_contrib *p_c2,
+ int w,
+ int incr5,
+ int incr3,
+ double ***p_c_S,
+ double ***p_c2_S){
+
+ unsigned int i;
+ int pc_size, j;
+ *p_c_S = (double **)space(sizeof(double *)*(n1+1));
+
+ for (i=1; i<=n1; i++){
+ pc_size = MIN2((w + incr5 + incr3), (int)n1);
+ (*p_c_S)[i] = (double *)space(sizeof(double) * (pc_size + 1));
+ for (j=0; j < pc_size; j++)
+ (*p_c_S)[i][j] = p_c->H[i][j] + p_c->I[i][j] + p_c->M[i][j] + p_c->E[i][j];
+ }
+
+ if(p_c2 != NULL){
+ (*p_c2_S) = (double **)space(sizeof(double *) * (n2 + 1));
+ for (i=1; i<=n2; i++){
+ pc_size = MIN2(w, (int)n2);
+ (*p_c2_S)[i] = (double *)space(sizeof(double) * (pc_size + 2));
+ for (j=0; j < pc_size; j++)
+ (*p_c2_S)[i][j] = p_c2->H[i][j] + p_c2->I[i][j] + p_c2->M[i][j] + p_c2->E[i][j];
+ }
+ }
+}
+
+/*------------------------------------------------------------------------*/
+/* s1 is the longer seq */
+PUBLIC interact *pf_interact( const char *s1,
+ const char *s2,
+ pu_contrib *p_c,
+ pu_contrib *p_c2,
+ int w,
+ char *cstruc,
+ int incr3,
+ int incr5){
+
+ int i, j, k,l,n1,n2,add_i5,add_i3,i_max,k_max, pc_size;
+ double temp, Z, rev_d, E, Z2,**p_c_S, **p_c2_S, int_scale;
+ FLT_OR_DBL ****qint_4, **qint_ik;
+ /* PRIVATE double **pint; array for pf_up() output */
+ interact *Int;
+ double G_min, G_is,Gi_min;
+ int gi,gj,gk,gl,ci,cj,ck,cl,prev_k,prev_l;
+ FLT_OR_DBL **int_ik;
+ double Z_int, temp_int, temppfs;
+ double const_scale, const_T;
+ constrain *cc = NULL; /* constrains for cofolding */
+ char *Seq, *i_long,*i_short,*pos=NULL; /* short seq appended to long one */
+ /* int ***pu_jl; */ /* positions of interaction in the short RNA */
+
+ G_min = G_is = Gi_min = 100.0;
+ gi = gj = gk = gl = ci = cj = ck = cl = 0;
+
+ n1 = (int) strlen(s1);
+ n2 = (int) strlen(s2);
+ prev_k = 1;
+ prev_l = n2;
+
+ i_long = (char *) space (sizeof(char)*(n1+1));
+ i_short = (char *) space (sizeof(char)*(n2+1));
+ Seq = (char *) space (sizeof(char)*(n1+n2+2));
+
+ strcpy(Seq,s1);
+ strcat(Seq,s2);
+
+ set_encoded_seq(s1, &S, &S1);
+ set_encoded_seq(s2, &SS, &SS2);
+
+ cc = get_ptypes(Seq,cstruc);
+
+ get_interact_arrays(n1, n2, p_c, p_c2, w, incr5, incr3, &p_c_S, &p_c2_S);
+
+ /*array for pf_up() output */
+ Int = (interact *) space(sizeof(interact)*1);
+ Int->Pi = (double *) space(sizeof(double)*(n1+2));
+ Int->Gi = (double *) space(sizeof(double)*(n1+2));
+
+ /* use a different scaling for pf_interact*/
+ scale_int(s2, s1, &int_scale);
+
+ /* set the global scale array and the global variable pf_scale to the
+ values used to scale the interaction, keep their former values !! */
+ temppfs = pf_scale;
+ pf_scale = int_scale;
+
+ /* in order to scale expLoopEnergy correctly call*/
+ scale_stru_pf_params((unsigned) n1);
+
+ qint_ik = (FLT_OR_DBL **) space(sizeof(FLT_OR_DBL *) * (n1+1));
+ for (i=1; i<=n1; i++) {
+ qint_ik[i] = (FLT_OR_DBL *) space(sizeof(FLT_OR_DBL) * (n1+1));
+ }
+/* int_ik */
+ int_ik = (FLT_OR_DBL **) space(sizeof(FLT_OR_DBL *) * (n1+1));
+ for (i=1; i<=n1; i++) {
+ int_ik[i] = (FLT_OR_DBL *) space(sizeof(FLT_OR_DBL) * (n1+1));
+ }
+ Z_int=0.;
+ /* Gint = ( -log(int_ik[gk][gi])-( ((int) w/2)*log(pf_scale)) )*((Pf->temperature+K0)*GASCONST/1000.0); */
+ const_scale = ((int) w/2)*log(pf_scale);
+ const_T = (Pf->kT/1000.0);
+ encode_seq(s1, s2);
+ /* static short *S~S1, *S1~SS1, *SS~S2, *SS2; */
+ for (i=0; i<=n1; i++) {
+ Int->Pi[i]=Int->Gi[i]=0.;
+ }
+ E=0.;
+ Z=0.;
+
+ if ( fold_constrained && cstruc != NULL) {
+ pos = strchr(cstruc,'|');
+ if(pos) {
+ ci=ck=cl=cj=0;
+ /* long seq & short seq
+ .........||..|||||....&....||||... w = maximal interaction length
+ ck ci cj cl */
+ strncpy(i_long,cstruc,n1);
+ i_long[n1] = '\0';
+ strncpy(i_short,&cstruc[n1],n2);
+ i_short[n2] ='\0';
+ pos = strchr(i_long,'|');
+ if(pos) ck = (int) (pos-i_long)+1; /* k */
+ pos = strrchr(i_long,'|');
+ if(pos) ci = (int) (pos-i_long)+1; /* i */
+ pos = strrchr(i_short,'|');
+ if(pos) cl = (int) (pos-i_short)+1; /* l */
+ pos = strchr(i_short,'|');
+ if(pos) cj = (int) (pos-i_short)+1; /* j */
+
+ if(ck > 0 && ci > 0 && ci-ck+1 > w) {
+ fprintf(stderr, "distance between constrains in longer seq, %d, larger than -w = %d",ci-ck+1,w);
+ nrerror("pf_interact: could not satisfy all constraints");
+ }
+ if(cj > 0 && cl > 0 && cl-cj+1 > w) {
+ fprintf(stderr, "distance between constrains in shorter seq, %d, larger than -w = %d",cl-cj+1,w);
+ nrerror("pf_interact: could not satisfy all constraints");
+ }
+ }
+
+ } else if ( fold_constrained && cstruc == NULL) {
+ nrerror("option -C selected, but no constrained structure given\n");
+ }
+ if(fold_constrained) pos = strchr(cstruc,'|');
+
+ /* qint_4[i][j][k][l] contribution that region (k-i) in seq1 (l=n1)
+ is paired to region (l-j) in seq 2(l=n2) that is
+ a region closed by bp k-l and bp i-j */
+ qint_4 = (FLT_OR_DBL ****) space(sizeof(FLT_OR_DBL ***) * (n1+1));
+
+ /* qint_4[i][j][k][l] */
+ for (i=1; i<=n1; i++) {
+ int end_k;
+ end_k = i-w;
+ if(fold_constrained && pos && ci) end_k= MAX2(i-w, ci-w);
+ /* '|' constrains for long sequence: index i from 1 to n1 (5' to 3')*/
+ /* interaction has to include 3' most '|' constrain, ci */
+ if(fold_constrained && pos && ci && i==1 && i<ci)
+ i= ci-w+1 > 1 ? ci-w+1 : 1;
+ /* interaction has to include 5' most '|' constrain, ck*/
+ if(fold_constrained && pos && ck && i > ck+w-1) break;
+
+ /* note: qint_4[i] will be freed before we allocate qint_4[i+1] */
+ qint_4[i] = (FLT_OR_DBL ***) space(sizeof(FLT_OR_DBL **) * (n2+1));
+ for (j=n2; j>0; j--) {
+ qint_4[i][j] = (FLT_OR_DBL **) space(sizeof(FLT_OR_DBL*) * (w+1));
+ for (k=0; k<=w; k++) {
+ qint_4[i][j][k] = (FLT_OR_DBL *) space(sizeof(FLT_OR_DBL) * (w+1));
+ }
+ }
+
+ prev_k=1;
+ for (j=n2; j>0; j--) {
+ int type, type2,end_l;
+ end_l = j+w;
+ if(fold_constrained && pos && ci) end_l= MIN2(cj+w,j+w);
+ /* '|' constrains for short sequence: index j from n2 to 1 (3' to 5')*/
+ /* interaction has to include 5' most '|' constrain, cj */
+ if(fold_constrained && pos && cj && j==n2 && j>cj)
+ j = cj+w-1 > n2 ? n2 : cj+w-1;
+ /* interaction has to include 3' most '|' constrain, cl*/
+ if(fold_constrained && pos && cl && j < cl-w+1) break;
+ type = cc->ptype[cc->indx[i]-(n1+j)];
+ qint_4[i][j][0][0] = type ? Pf->expDuplexInit : 0;
+
+ if (!type) continue;
+ qint_4[i][j][0][0] *= exp_E_ExtLoop(type, (i>1) ? S1[i-1] : -1, (j<n2) ? SS2[j+1] : -1, Pf);
+
+ rev_d = exp_E_ExtLoop(rtype[type], (j>1) ? SS2[j-1] : -1, (i<n1) ? S1[i+1] : -1, Pf);
+
+ /* add inc5 and incr3 */
+ if((i-incr5) > 0 ) add_i5=i-incr5;
+ else add_i5=1;
+ add_i3=incr3;
+ pc_size = MIN2((w+incr3+incr5),n1);
+ if(incr3 < pc_size) add_i3=incr3;
+ else add_i3=pc_size-1;
+
+ /* only one bp (no interior loop) */
+ if(p_c2 == NULL) {/* consider only structure of longer seq. */
+ qint_ik[i][i]+=qint_4[i][j][0][0]*rev_d*p_c_S[add_i5][add_i3]*scale[((int) w/2)];
+ Z+=qint_4[i][j][0][0]*rev_d*p_c_S[add_i5][add_i3]*scale[((int) w/2)];
+ } else {/* consider structures of both seqs. */
+ qint_ik[i][i]+=qint_4[i][j][0][0]*rev_d*p_c_S[add_i5][add_i3]*p_c2_S[j][0]*scale[((int) w/2)];
+ Z+=qint_4[i][j][0][0]*rev_d*p_c_S[add_i5][add_i3]*p_c2_S[j][0]*scale[((int) w/2)];
+ }
+
+/* int_ik */
+ /* check deltaG_ges = deltaG_int + deltaG_unstr; */
+ int_ik[i][i]+=qint_4[i][j][0][0]*rev_d*scale[((int) w/2)];
+ Z_int+=qint_4[i][j][0][0]*rev_d*scale[((int) w/2)];
+ temp_int=0.;
+
+ temp=0.;
+ prev_l = n2;
+ for (k=i-1; k>end_k && k>0; k--) {
+ if (fold_constrained && pos && cstruc[k-1] == '|' && k > prev_k)
+ prev_k=k;
+ for (l=j+1; l< end_l && l<=n2; l++) {
+ int a,b,ia,ib,isw;
+ double scalew, tt, intt;
+
+ type2 = cc->ptype[cc->indx[k]-(n1+l)];
+ /* '|' : l HAS TO be paired: not pair (k,x) where x>l allowed */
+ if(fold_constrained && pos && cstruc[n1+l-1] == '|' && l < prev_l)
+ prev_l=l; /*break*/
+ if(fold_constrained && pos && (k<=ck || i>=ci) && !type2) continue;
+ if(fold_constrained && pos && ((cstruc[k-1] == '|') || (cstruc[n1+l-1] == '|')) && !type2) break;
+
+ if (!type2) continue;
+ /* to save memory keep only qint_4[i-w...i][][][] in memory
+ use indices qint_4[i][j][a={0,1,...,w-1}][b={0,1,...,w-1}] */
+ a=i-k;/* k -> a from 1...w-1*/
+ b=l-j;/* l -> b from 1...w-1 */
+
+ /* scale everything to w/2 */
+ isw = ((int) w/2);
+ if ((a+b) < isw ){
+ scalew = ( scale[isw - (a+b)] );
+ } else if ( (a+b) > isw ) {
+ scalew = 1/( scale[(a+b) - isw] );
+ } else {
+ scalew = 1;
+ }
+
+ if (i-k+l-j-2<=MAXLOOP) {
+ if(k >= prev_k && l <= prev_l) { /* don't violate constrains */
+ E = exp_E_IntLoop(i-k-1,l-j-1, type2, rtype[type],
+ S1[k+1], SS2[l-1], S1[i-1], SS2[j+1], Pf) *
+ scale[i-k+l-j]; /* add *scale[u1+u2+2] */
+
+ qint_4[i][j][a][b] += ( qint_4[k][l][0][0]*E);
+
+ /* use ia and ib to go from a....w-1 and from b....w-1 */
+ ia=ib=1;
+ while((a+ia)<w && i-(a+ia)>=1 && (b+ib)<w && (j+b+ib)<=n2) {
+ int iaa,ibb;
+
+ qint_4[i][j][a+ia][b+ib] += qint_4[k][l][ia][ib]*E;
+
+ iaa=ia+1;
+ while(a+iaa<w && i-(a+iaa)>=1) {
+ qint_4[i][j][a+iaa][b+ib] += qint_4[k][l][iaa][ib]*E;
+ ++iaa;
+ }
+
+ ibb=ib+1;
+ while( (b+ibb)<w && (j+b+ibb)<=n2 ) {
+ qint_4[i][j][a+ia][b+ibb] += qint_4[k][l][ia][ibb]*E;
+ ++ibb;
+ }
+ ++ia;
+ ++ib;
+ }
+ }
+ }
+ /* '|' constrain in long sequence */
+ /* collect interactions starting before 5' most '|' constrain */
+ if ( fold_constrained && pos && ci && i < ci) continue;
+ /* collect interactions ending after 3' most '|' constrain*/
+ if ( fold_constrained && pos && ck && k > ck) continue;
+ /* '|' constrain in short sequence */
+ /* collect interactions starting before 5' most '|' constrain */
+ if ( fold_constrained && pos && cj && j > cj) continue;
+ /* collect interactions ending after 3' most '|' constrain*/
+ if ( fold_constrained && pos && cl && l < cl) continue;
+
+ /* scale everything to w/2*/
+ /* qint_ik[k][i] all interactions where k and i both are paired */
+ /* substract incr5 from k */
+ if(k-incr5 > 0) add_i5=k-incr5;
+ else add_i5=1;
+ /* add incr3 to i */
+ pc_size = MIN2((w+incr3+incr5),n1);
+ if(i-k+incr3 < pc_size) add_i3=i-k+incr3;
+ else add_i3=pc_size-1;
+
+ if(p_c2 == NULL) {/* consider only structure of longer seq. */
+ tt = qint_4[i][j][a][b]*p_c_S[add_i5][add_i3]*scalew*rev_d;
+ } else { /* consider structures of both seqs. */
+ tt = qint_4[i][j][a][b]*p_c_S[add_i5][add_i3]*p_c2_S[j][b]*scalew*rev_d;
+ }
+ temp+= tt;
+ qint_ik[k][i]+= tt;
+ /* int_ik */
+ /* check deltaG_ges = deltaG_int + deltaG_unstr; */
+ intt = qint_4[i][j][a][b]*scalew*rev_d;
+ temp_int += intt;
+ int_ik[k][i]+= intt;
+ G_is = (-log(tt)-const_scale)*(const_T);
+ if (G_is < G_min || EQUAL(G_is,G_min)) {
+ G_min = G_is;
+ Gi_min =(-log(intt)-const_scale)*(const_T);
+ gi=i;
+ gj=j;
+ gk=k;
+ gl=l;
+ }
+ }
+ }
+ Z+=temp;
+ /* int_ik */
+ Z_int+=temp_int;
+ }
+
+ /* free qint_4 values not needed any more */
+ if(i > w) {
+ int bla;
+ bla=i-w;
+ if (fold_constrained && pos && ci && i-w < ci-w+1) continue;
+ if (fold_constrained && pos && ci) bla = MAX2(ci-w+1,i-w);
+ for (j=n2; j>0; j--) {
+ for (k=0; k<=w; k++){
+ free(qint_4[bla][j][k]);
+ }
+ free(qint_4[bla][j]);
+ }
+ free(qint_4[bla]);
+ qint_4[bla] = NULL;
+ }
+ }
+
+
+ Z2=0.0;
+ i_max = k_max = 0;
+ for (i=1; i<=n1; i++) {
+ for (k=i; k<=n1 && k<i+w; k++) {
+ Z2+=qint_ik[i][k];
+ for(l=i;l<=k;l++) {
+ /* Int->Pi[l]: prob that position l is within a paired region */
+ /* qint_ik[i][k] as well as Z are scaled to scale[((int) w/2) */
+ Int->Pi[l]+=qint_ik[i][k]/Z;
+ /* Int->Gi[l]: minimal delta G at position [l] */
+ Int->Gi[l]=MIN2(Int->Gi[l],
+ ( -log(qint_ik[i][k])-( ((int) w/2)*log(pf_scale)) )*
+ (Pf->kT/1000.0) );
+ }
+ }
+ }
+ if(n1 > w){
+ int start_i,end_i;
+ start_i = n1-w+1;
+ end_i=n1;
+ if (fold_constrained && pos && ci) {
+ /* a break in the k loop might result in unfreed values */
+ start_i = ci-w+1 < n1-w+1 ? ci-w+1 : n1-w+1;
+ start_i = start_i > 0 ? start_i : 1;
+ /* start_i = ck; */
+ end_i = ck+w-1 > n1 ? n1 : ck+w-1;
+ }
+ for (i=start_i; i<=end_i; i++) {
+ if(qint_4[i] == NULL ) continue;
+ for (j=n2; j>0; j--) {
+ for (k=0; k<=w; k++) {
+ free(qint_4[i][j][k]);
+ }
+ free(qint_4[i][j]);
+ }
+ free(qint_4[i]);
+ }
+ free(qint_4);
+ } else {
+ int start_i,end_i;
+ start_i = 1;
+ end_i=n1;
+ if (fold_constrained && pos) {
+ start_i = ci-w+1 > 0 ? ci-w+1 : 1;
+ end_i = ck+w-1 > n1 ? n1 : ck+w-1;
+ }
+
+ for (i=start_i; i<=end_i; i++) {
+ for (j=n2; j>0; j--) {
+ for (k=0; k<=w; k++) {
+ free(qint_4[i][j][k]);
+ }
+ free(qint_4[i][j]);
+ }
+ free(qint_4[i]);
+ }
+ free(qint_4);
+ }
+ if(fold_constrained && (gi==0 || gk==0 || gl==0 || gj==0)) {
+ nrerror("pf_interact: could not satisfy all constraints");
+ }
+ /* fill structure interact */
+ Int->length = n1;
+ Int->i = gi;
+ Int->j = gj;
+ Int->k = gk;
+ Int->l = gl;
+ Int->Gikjl = G_min;
+ Int->Gikjl_wo = Gi_min;
+
+ free(i_long);
+ free(i_short);
+
+ for (i=1; i<=n1; i++) {
+ free(int_ik[i]);
+ }
+ free(int_ik);
+ for (i=1; i<=n1; i++) {
+ free(qint_ik[i]);
+ }
+ free(qint_ik);
+
+ /* reset the global variables pf_scale and scale to their original values */
+ pf_scale = temppfs;/* reset pf_scale */
+ scale_stru_pf_params((unsigned) n1);/* reset the scale array */
+ free_pf_arrays(); /* for arrays for pf_fold(...) */
+
+ if(expMLbase != NULL) {
+ free(expMLbase);
+ expMLbase = NULL;
+ }
+ if(scale != NULL) {
+ free(scale);
+ scale = NULL;
+ }
+ for (i=1; i<=n1; i++) {
+ free(p_c_S[i]);
+ }
+ free(p_c_S);
+ if(p_c2 != NULL) {
+ for (i=1; i<=n2; i++) {
+ free(p_c2_S[i]);
+ }
+ free(p_c2_S);
+ }
+ free(Seq);
+ free(cc->indx);
+ free(cc->ptype);
+ free(cc);
+ return(Int);
+}
+/*------------------------------------------------------------------------*/
+/* use an extra scale for pf_interact, here sl is the longer sequence */
+PRIVATE void scale_int(const char *s, const char *sl, double *sc_int){
+ int n,nl,l_scales;
+ duplexT mfe;
+ double kT;
+
+ n = strlen(s);
+ nl = strlen(sl);
+
+ expMLbase = (FLT_OR_DBL *) space(sizeof(FLT_OR_DBL)*(nl+1));
+ scale = (FLT_OR_DBL *) space(sizeof(FLT_OR_DBL)*((nl+1)*2));
+
+ /* use RNA duplex to get a realistic estimate for the best possible
+ interaction energy between the short RNA s and its target sl */
+ mfe = duplexfold(s,sl);
+
+ kT = Pf->kT/1000.0; /* in Kcal */
+
+ /* sc_int is similar to pf_scale: i.e. one time the scale */
+ *sc_int = exp(-(mfe.energy)/kT/n);
+
+ /* free the structure returned by duplexfold */
+ free(mfe.structure);
+}
+
+/*----------------------------------------------------------------------*/
+/* init_pf_two(n) :gets the arrays, that you need, from part_func.c */
+/* get_pf_arrays(&S, &S1, &ptype, &qb, &qm, &q1k, &qln);*/
+/* init_pf_fold(), update_pf_params, encode_char(), make_ptypes() are called by pf_fold() */
+PRIVATE void init_pf_two(int length){
+#ifdef SUN4
+ nonstandard_arithmetic();
+#else
+#ifdef HP9
+ fpsetfastmode(1);
+#endif
+#endif
+ make_pair_matrix();
+
+ /* gets the arrays, that we need, from part_func.c */
+ if(!get_pf_arrays(&S, &S1, &ptype, &qb, &qm, &q1k, &qln))
+ nrerror("init_pf_two: pf_fold() has to be called before calling pf_unstru()\n");
+ /* get a pointer to the base pair probs */
+ probs = export_bppm();
+
+ scale_stru_pf_params((unsigned) length);
+
+ init_length=length;
+ if(init_temp != Pf->temperature)
+ nrerror("init_pf_two: inconsistency with temperature");
+}
+
+PRIVATE void get_up_arrays(unsigned int length){
+ unsigned int l1 = length + 1;
+ unsigned int l2 = length + 2;
+ prpr = (FLT_OR_DBL *)space(sizeof(FLT_OR_DBL) * ((l1*l2)>>1));
+ expMLbase = (FLT_OR_DBL *)space(sizeof(FLT_OR_DBL) * l1);
+ scale = (FLT_OR_DBL *)space(sizeof(FLT_OR_DBL) * l1);
+ qqm2 = (double *) space(sizeof(double) * l2);
+ qq_1m2 = (double *) space(sizeof(double) * l2);
+ qqm = (double *) space(sizeof(double) * l2);
+ qqm1 = (double *) space(sizeof(double) * l2);
+ my_iindx = get_iindx(length);
+}
+
+PRIVATE void free_up_arrays(void){
+ if(prpr != NULL){ free(prpr); prpr = NULL;}
+ if(expMLbase != NULL){ free(expMLbase); expMLbase = NULL;}
+ if(scale != NULL){ free(scale); scale = NULL;}
+ if(qqm != NULL){ free(qqm); qqm = NULL;}
+ if(qqm1 != NULL){ free(qqm1); qqm1 = NULL;}
+ if(qqm2 != NULL){ free(qqm2); qqm2 = NULL;}
+ if(qq_1m2 != NULL){ free(qq_1m2); qq_1m2 = NULL;}
+ if(my_iindx != NULL){ free(my_iindx); my_iindx = NULL;}
+}
+
+PUBLIC void free_interact(interact *pin) {
+ if(S != NULL && pin != NULL){
+ free(S);
+ S=NULL;
+ }
+ if(S1 != NULL && pin != NULL){
+ free(S1);
+ S1=NULL;
+ }
+ if(pin != NULL){
+ free(pin->Pi);
+ free(pin->Gi);
+ free(pin);
+ pin=NULL;
+ }
+}
+/*---------------------------------------------------------------------------*/
+
+PRIVATE void encode_seq(const char *s1, const char *s2) {
+ unsigned int i,l;
+
+ l = strlen(s1);
+ /* S and S1 are freed by free_pf_arrays(); ! */
+ S = (short *) space(sizeof(short)*(l+1));
+ S1= (short *) space(sizeof(short)*(l+1));
+ /* S1 exists only for the special X K and I bases and energy_set!=0 */
+ S[0] = l;
+ for (i=1; i<=l; i++) { /* make numerical encoding of sequence */
+ S[i]= (short) encode_char(toupper(s1[i-1]));
+ S1[i] = alias[S[i]]; /* for mismatches of nostandard bases */
+ }
+ if(s2 != NULL) {
+ l = strlen(s2);
+ /* SS2 exists only for the special X K and I bases and energy_set!=0 */
+ SS[0] = l;
+ for (i=1; i<=l; i++) { /* make numerical encoding of sequence */
+ SS[i]= (short) encode_char(toupper(s2[i-1]));
+ SS2[i] = alias[SS[i]]; /* for mismatches of nostandard bases */
+ }
+ }
+}
+
+/*-------------------------------------------------------------------------*/
+ /* scale energy parameters and pre-calculate Boltzmann weights:
+ most of this is done in structure Pf see params.c,h (function:
+ get_scaled_pf_parameters(), only arrays scale and expMLbase are handled here*/
+PRIVATE void scale_stru_pf_params(unsigned int length)
+{
+ unsigned int i;
+ double kT;
+
+
+ /* Do this only at the first call for get_scaled_pf_parameters()
+ and/or if temperature has changed*/
+ if(init_temp != temperature) {
+ if(Pf) free(Pf);
+ Pf=get_scaled_pf_parameters();
+ }
+
+ init_temp = Pf->temperature;
+
+ kT = Pf->kT; /* kT in cal/mol */
+
+ /* scaling factors (to avoid overflows) */
+ if (pf_scale == -1) { /* mean energy for random sequences: 184.3*length cal */
+ pf_scale = exp(-(-185+(Pf->temperature-37.)*7.27)/kT);
+ if (pf_scale<1) pf_scale=1;
+ }
+ scale[0] = 1.;
+ scale[1] = 1./pf_scale;
+ expMLbase[0] = 1;
+ expMLbase[1] = Pf->expMLbase/pf_scale;
+ for (i=2; i<=length; i++) {
+ scale[i] = scale[i/2]*scale[i-(i/2)];
+ expMLbase[i] = pow(Pf->expMLbase, (double)i) * scale[i];
+ }
+}
+/*-------------------------------------------------------------------------*/
+/* make a results structure containing all u-values & the header */
+PUBLIC pu_out *get_u_vals(pu_contrib *p_c, int **unpaired_values, char *select_contrib) {
+ int i, j, k, l, g,ws,num_u_vals,unstr,count,contribs,size,w,len;
+ int S,E,H,I,M;
+ int off_S, off_E, off_H, off_I, off_M;
+ /* double **p_cont,**p_cont_sh, dG_u; p_u AND its contributions */
+ pu_out* u_results;
+
+ len = p_c->length;
+
+ /* number of different -u values */
+ for (num_u_vals = 0, i = 1; i <= unpaired_values[0][0]; i++) {
+ j = unpaired_values[i][0];
+ do num_u_vals++; while(++j <= unpaired_values[i][1]);
+ }
+ /* check which contributions ([-c "SHIME"] ) are desired by the user,
+ set the offset for each contribution */
+ contribs = 0;
+ S = E = H = I = M = 0;
+ off_S = off_E = off_H = off_I = off_M = 0;
+ if(strchr(select_contrib, 'S')) {
+ S=1;
+ off_S = contribs;
+ ++contribs;
+ }
+ if(strchr(select_contrib, 'E')) {
+ E=1;
+ off_E = contribs;
+ ++contribs;
+ }
+ if(strchr(select_contrib, 'H')) {
+ H=1;
+ off_H = contribs;
+ ++contribs;
+ }
+ if(strchr(select_contrib, 'I')) {
+ I=1;
+ off_I = contribs;
+ ++contribs;
+ }
+ if(strchr(select_contrib, 'M')) {
+ M=1;
+ off_M = contribs;
+ ++contribs;
+ }
+
+ if(contribs > 5) {
+ nrerror("get_u_vals: error with contribs!");
+ }
+ /* allocate the results structure */
+ u_results = (pu_out *) space(1*sizeof(pu_out));
+ u_results->len = len; /* sequence length */
+ /*num_u_vals differnet -u values, contribs [-c "SHIME"] */
+ u_results->u_vals = num_u_vals;
+ u_results->contribs = contribs;
+ /* add 1 column for position within the sequence and
+ add 1 column for the free energy of interaction values */
+ /* header e.g. u3I (contribution for u3 interior loops */
+ size = 1 + (num_u_vals*contribs) + 1;
+ u_results->header = (char **) space((size+1)*sizeof(char*));
+ for(i=0;i<(size+1);i++){
+ u_results->header[i] = (char *) space(10*sizeof(char));
+ }
+ /* different free energies for all -u and -c combinations */
+ u_results->u_values = (double**) space((size+1) *sizeof(double*));
+ for(i=0;i<(size+1);i++){
+ /* position within the sequence */
+ u_results->u_values[i] = (double*) space((len+3)*sizeof(double));
+ }
+ /* write the position within the sequence in the u_results array
+ at column zerro */
+ sprintf(u_results->header[0],"pos");
+ for(i=0;i<=len;i++){
+ /* add the position*/
+ u_results->u_values[0][i] = i;
+ }
+ /* go over the different -u values, u_vals[] listy of different -u values*/
+ for (count = k = 1; k <= unpaired_values[0][0]; k++) {
+ l = unpaired_values[k][0];
+ do{
+ int offset; /* offset for the respective -u value (depents on the number
+ of the -u value and on the numbers of contribs */
+
+ offset = ((count - 1) * contribs) + 1; /* first colum is the position */
+ /* set the current value of -u : here we call it w */
+ w = l; /* set w to the actual -u value */
+ if(w > len) break; /* corr caro */
+ /* make the header - look which contribitions are wanted */
+ if(S) sprintf(u_results->header[offset+off_S],"u%dS",w);
+ if(E) sprintf(u_results->header[offset+off_E],"u%dE",w);
+ if(H) sprintf(u_results->header[offset+off_H],"u%dH",w);
+ if(I) sprintf(u_results->header[offset+off_I],"u%dI",w);
+ if(M) sprintf(u_results->header[offset+off_M],"u%dM",w);
+
+ if(p_c != NULL) {
+ for (i=1; i<=len; i++) { /* for each position */
+ /* w goes form j to i (intervall end at i) */
+ for (j=i; j < MIN2((i+w),len+1); j++) { /* for each -u value < w
+ this is not necessay ->
+ calculate j from i and w
+ : (j-i+1) == w */
+ double blubb;
+ /* if (j-i+1) == w we have the -u = w value wanted */
+ if( (j-i+1) == w && i+w-1 <= len) {
+ blubb = p_c->H[i][j-i]+p_c->I[i][j-i]+p_c->M[i][j-i]+p_c->E[i][j-i];
+
+ /* printf("len %d blubb %.3f \n",len, blubb); */
+ if(S) u_results->u_values[offset+off_S][i+w-1]+=blubb;
+ if(E) u_results->u_values[offset+off_E][i+w-1]+=p_c->E[i][j-i];
+ if(H) u_results->u_values[offset+off_H][i+w-1]+=p_c->H[i][j-i];
+ if(I) u_results->u_values[offset+off_I][i+w-1]+=p_c->I[i][j-i];
+ if(M) u_results->u_values[offset+off_M][i+w-1]+=p_c->M[i][j-i];
+
+ }
+ if(i<w && (j-i+1) != w && i+w-1 > len && i+w-1 < len+3) {
+ if(S) u_results->u_values[offset+off_S][i+w-1]=-1;
+ if(E) u_results->u_values[offset+off_E][i+w-1]=-1;
+ if(H) u_results->u_values[offset+off_H][i+w-1]=-1;
+ if(I) u_results->u_values[offset+off_I][i+w-1]=-1;
+ if(M) u_results->u_values[offset+off_M][i+w-1]=-1;
+ }
+ }
+ }
+ } else return(NULL); /* error */
+ count++;
+ } while(++l <= unpaired_values[k][1]);
+ }
+ return(u_results); /*success*/
+}
+/* plot the results structure */
+/* when plotting the results for the target seq we add a header */
+/* when plotting the results for the interaction partner u want no header,
+ set s1 to NULL to avoid plotting the header */
+/* currently we plot the free energies to a file: the probability of
+ being unpaired for region [i,j], p_u[i,j], is related to the free
+ energy to open region [i,j], dG_u[i,j] by:
+ dG_u[i,j] = -log(p_u[i,j])*(temperature+K0)*GASCONST/1000.0; */
+PUBLIC int plot_free_pu_out(pu_out* res, interact *pint, char *ofile, char *head) {
+ int size,s,i,len;
+ double dG_u;
+ char nan[4], *time, dg[11];
+ FILE *wastl;
+ double kT = Pf->kT;
+ wastl = fopen(ofile,"a");
+ if (wastl==NULL) {
+ fprintf(stderr, "p_cont: can't open %s for Up_plot\n", ofile);
+ return(0);
+ }
+ sprintf(dg,"dG");
+
+ /* printf("T=%.16f \n(temperature+K0)*GASCONST/1000.0 = %.16f\n",temperature,(temperature+K0)*GASCONST/1000.0); */
+
+ /* write the header of the output file: */
+ /* # timestamp commandlineaufruf */
+ /* # length and name of first sequence (target)
+ /* # first seq */
+ /* # length and name of second sequence (interaction partner) */
+ /* # second seq */
+ /* the next line is the output for the target: colums
+ position in target | dG_unpaired values for target | interaction energy */
+ /* # pos u1S u1H dg */
+ /* values for target */
+ /* if -b was choosen: the next lines are the dG_unpaired values for
+ the interaction partner */
+ /* # pos u1S u1H */
+ /* values for the interaction partner */
+
+ /* print header, if nh is zerro */
+ if(head){
+ time = time_stamp();
+ fprintf(wastl,"# %s\n", time);
+ fprintf(wastl,"%s\n",head);
+ }
+ fprintf(wastl,"# ");
+ /* } else { fprintf(wastl," "); } close if before */
+ len = res->len;
+ size = res->u_vals * res->contribs;
+
+ sprintf(nan,"NA");
+ nan[2] = '\0';
+
+ for(i=0;i<=len; i++) {
+ for(s=0;s<=size+1;s++) { /* that is for different contribution */
+ if ( i== 0 && s > size && pint != NULL)
+ fprintf(wastl,"%8s ",dg);
+ if(i != 0) {
+ if(s>0 && s<=size) {
+ if(res->u_values[s][i] > 0.0) {
+ dG_u = -log(res->u_values[s][i])*kT/1000.0;
+ fprintf(wastl,"%8.3f ",dG_u);
+ } else { /* no p_u value was defined print nan*/
+ fprintf(wastl,"%8s ",nan);
+ }
+
+ } else if (s > size && pint != NULL) {
+ fprintf(wastl,"%8.3f ",pint->Gi[i]);
+ } else if (s == 0) {
+ fprintf(wastl,"%8.0f ",res->u_values[s][i]);
+ }
+ } else {
+ if(s>1) {
+ fprintf(wastl,"%8s ",res->header[s]);
+ } else {
+ fprintf(wastl,"%7s ",res->header[s]);
+ }
+ }
+ }
+ fprintf(wastl,"\n");
+ }
+ fclose(wastl);
+ /*free pu_out* res */
+ if(res != NULL) {
+ for(i=0;i<=(size+2);i++) {
+ free(res->u_values[i]);
+ free(res->header[i]);
+ }
+ free(res->u_values);
+ free(res->header);
+ free(res);
+ res = NULL;
+ }
+
+ return(1); /* success */
+}
+
+PUBLIC int Up_plot(pu_contrib *p_c, pu_contrib *p_c_sh, interact *pint, char *ofile, int **unpaired_values, char *select_contrib, char *head, unsigned int mode) {
+ pu_out *dada;
+ int ret;
+ /* check what case we have */
+
+ /* upmode = 1 only one seq */
+ /* if(p_c != NULL && pint == NULL) { */
+ if(mode & RNA_UP_MODE_1){
+ dada = get_u_vals(p_c,unpaired_values,select_contrib);
+ ret = plot_free_pu_out(dada,NULL,ofile,head);
+
+ /* upmode > 1 cofolding */
+ /* } else if (p_c != NULL && pint != NULL) { */
+ } else if(mode & RNA_UP_MODE_2) {
+ dada = get_u_vals(p_c,unpaired_values,select_contrib);
+ ret = plot_free_pu_out(dada,pint,ofile,head);
+
+ /* upmode = 3 cofolding*/
+ /* } else if (p_c == NULL && p_c_sh != NULL) { */
+ }
+ if(mode & RNA_UP_MODE_3) {
+ dada = get_u_vals(p_c,unpaired_values, select_contrib);
+ ret = plot_free_pu_out(dada, pint, ofile, head);
+
+ dada = get_u_vals(p_c_sh, unpaired_values, select_contrib);
+ ret = plot_free_pu_out(dada,NULL,ofile, NULL);
+ }
+ return(ret);
+}
+
+/*-------------------------------------------------------------------------*/
+/* copy from part_func_co.c */
+PRIVATE constrain *get_ptypes(char *Seq, const char *structure) {
+ int n,i,j,k,l, length;
+ constrain *con;
+ short *s, *s1;
+
+ length = strlen(Seq);
+ make_pair_matrix();
+ con = (constrain *) space(sizeof(constrain));
+ con->indx = (int *) space(sizeof(int)*(length+1));
+ for (i=1; i<=length; i++) {
+ con->indx[i] = ((length+1-i)*(length-i))/2 +length+1;
+ }
+ con->ptype = (char *) space(sizeof(char)*((length+1)*(length+2)/2));
+
+ set_encoded_seq((const char *)Seq, &s, &s1);
+
+ n=s[0];
+ for (k=1; k<=n-CO_TURN-1; k++)
+ for (l=1; l<=2; l++) {
+ int type,ntype=0,otype=0;
+ i=k; j = i+CO_TURN+l; if (j>n) continue;
+ type = pair[s[i]][s[j]];
+ while ((i>=1)&&(j<=n)) {
+ if ((i>1)&&(j<n)) ntype = pair[s[i-1]][s[j+1]];
+ if (noLonelyPairs && (!otype) && (!ntype))
+ type = 0; /* i.j can only form isolated pairs */
+ con->ptype[con->indx[i]-j] = (char) type;
+ otype = type;
+ type = ntype;
+ i--; j++;
+ }
+ }
+
+ if (fold_constrained&&(structure!=NULL)) {
+ int hx, *stack;
+ char type;
+ stack = (int *) space(sizeof(int)*(n+1));
+ for(hx=0, j=1; j<=n; j++) {
+ switch (structure[j-1]) {
+ case 'x': /* can't pair */
+ for (l=1; l<j-CO_TURN; l++) con->ptype[con->indx[l]-j] = 0;
+ for (l=j+CO_TURN+1; l<=n; l++) con->ptype[con->indx[j]-l] = 0;
+ break;
+ case '(':
+ stack[hx++]=j;
+ /* fallthrough */
+ case '<': /* pairs upstream */
+ break;
+ case ')':
+ if (hx<=0) {
+ fprintf(stderr, "%s\n", structure);
+ nrerror("1. unbalanced brackets in constraints");
+ }
+ i = stack[--hx];
+ type = con->ptype[con->indx[i]-j];
+ /* don't allow pairs i<k<j<l */
+ for (k=i; k<=j; k++)
+ for (l=j; l<=n; l++) con->ptype[con->indx[k]-l] = 0;
+ /* don't allow pairs k<i<l<j */
+ for (k=1; k<=i; k++)
+ for (l=i; l<=j; l++) con->ptype[con->indx[k]-l] = 0;
+ con->ptype[con->indx[i]-j] = (type==0)?7:type;
+ case '>': /* pairs downstream */
+ break;
+ }
+ }
+ if (hx!=0) {
+ fprintf(stderr, "%s\n", structure);
+ nrerror("2. unbalanced brackets in constraint string");
+ }
+ free(stack);
+ }
+ free(s);
+ free(s1);
+ return con;
+}
+PRIVATE void set_encoded_seq(const char *sequence, short **S, short **S1){
+ unsigned int i,l;
+ l = strlen(sequence);
+ if(S!= NULL){
+ *S = (short *)space(sizeof(short) * (l + 2));
+ for(i=1; i<=l; i++) /* make numerical encoding of sequence */
+ (*S)[i]= (short) encode_char(toupper(sequence[i-1]));
+ (*S)[l+1] = (*S)[1];
+ (*S)[0] = (short) l;
+ }
+ /* S1 exists only for the special X K and I bases and energy_set!=0 */
+ if(S1 != NULL){
+ *S1 = (short *)space(sizeof(short) * (l + 2));
+ for(i=1; i<=l; i++) /* make numerical encoding of sequence */
+ (*S1)[i] = alias[(short) encode_char(toupper(sequence[i-1]))]; /* for mismatches of nostandard bases */
+ /* for circular folding add first base at position n+1 and last base at position 0 in S1 */
+ (*S1)[l+1] = (*S1)[1];
+ (*S1)[0] = (*S1)[l];
+ }
+}
git://source.jalview.org / jabaws.git / blobdiff