PUBLIC int
move_standard(char *seq,
char *struc,
enum MOVE_TYPE type,
int verbosity_level,
int shifts,
int noLP){
make_pair_matrix();
short int *s0 = encode_sequence(seq, 0);
short int *s1 = encode_sequence(seq, 1);
short int *str = make_pair_table(struc);
int energy = 0;
switch (type){
case GRADIENT: energy = move_gradient(seq, str, s0, s1, verbosity_level, shifts, noLP); break;
case FIRST: energy = move_first(seq, str, s0, s1, verbosity_level, shifts, noLP); break;
case ADAPTIVE: energy = move_adaptive(seq, str, s0, s1, verbosity_level); break;
}
int i=1;
for (; i<=str[0]; i++) {
if (str[i]==0) struc[i-1]='.';
else if (str[i]>str[str[i]]) struc[i-1]='(';
else struc[i-1]=')';
}
free(s0);
free(s1);
free(str);
return energy;
}
bool is_legal_pair_content( char* seq, char* secstr, int ip )
{
if( cnf->max_solid_GCs <= 0 ) return true;
char* mark = strdup( secstr );
short* pt = make_pair_table( secstr );
int o;
for( o = 1; o <= strlen( cnf->hairpin_ss ); o++ ) {
int i = ip + o;
if( pt[i] < i ) continue;
mark[i-1] = ((seq[i-1] ^ seq[pt[i]-1] ^ 'G' ^ 'C')==0)? 'X' : '-';
}
bool ok = true;
char* p = mark;
while( (*p) && ok ) {
p += strcspn( p, "X" );
if( *p ) {
int n = strspn( p, "X" );
if( n > cnf->max_solid_GCs ) ok = false;
p += n;
}
}
free( pt );
free( mark );
return ok;
}
int main (int argc, char *argv[])
{
char *line, *sequence, *structure;
int i, j, length;
float energy;
short *ptable;
char *pair_seen;
int *iindx;
if (argc>1) help();
line = get_line (stdin);
if (*line=='>') {
free (line);
line = get_line (stdin);
}
/* read the sequence */
sequence = (char *) space (sizeof (char) * (strlen(line)+1));
(void) sscanf(line, "%s", sequence);
free (line);
length = (int) strlen(sequence);
iindx = (int *) space(sizeof(int)*(length+1));
for (i=1; i<=length; i++)
iindx[i] = ((length+1-i)*(length-i))/2 +length+1;
pair_seen = (char *) space(((length+1)*(length+2))/2 * sizeof(char));
structure = (char *) space (sizeof (char) * (length+1));
/* get of suboptimal structures */
while ((line = get_line (stdin))) {
int r, popt;
r = sscanf(line, "%s %f", structure, &energy);
free(line);
if (r==0) continue;
ptable = make_pair_table(structure);
popt = 0;
for (i = 0; i < length; i++) {
if ((j=ptable[i])>i)
if (pair_seen[iindx[i]-j]==0) {
pair_seen[iindx[i]-j]=(char) 1;
if (popt==0) {
printf("%s %6.2f", structure, energy);
popt = 1;
}
printf(" (%d,%d)", i,j);
}
}
if (popt==1) printf("\n");
free(ptable);
}
free(sequence); free(structure); free(iindx); free(pair_seen);
return 0;
}
PRIVATE void print_aliout(char **AS, plist *pl, double threshold, int n_seq, char * mfe, FILE *aliout) {
int i, j, k, n, num_p=0, max_p = 64;
pair_info *pi;
double *duck, p;
short *ptable;
for (n=0; pl[n].i>0; n++);
max_p = 64; pi = space(max_p*sizeof(pair_info));
duck = (double *) space((strlen(mfe)+1)*sizeof(double));
ptable = make_pair_table(mfe);
for (k=0; k<n; k++) {
int s, type;
p = pl[k].p; i=pl[k].i; j = pl[k].j;
duck[i] -= p * log(p);
duck[j] -= p * log(p);
if (p<threshold) continue;
pi[num_p].i = i;
pi[num_p].j = j;
pi[num_p].p = p;
pi[num_p].ent = duck[i]+duck[j]-p*log(p);
for (type=0; type<8; type++) pi[num_p].bp[type]=0;
for (s=0; s<n_seq; s++) {
int a,b;
a=encode_char(toupper(AS[s][i-1]));
b=encode_char(toupper(AS[s][j-1]));
type = pair[a][b];
if ((AS[s][i-1] == '-')||(AS[s][j-1] == '-')) type = 7;
if ((AS[s][i-1] == '~')||(AS[s][j-1] == '~')) type = 7;
pi[num_p].bp[type]++;
pi[num_p].comp = (ptable[i] == j) ? 1:0;
}
num_p++;
if (num_p>=max_p) {
max_p *= 2;
pi = xrealloc(pi, max_p * sizeof(pair_info));
}
}
free(duck);
pi[num_p].i=0;
qsort(pi, num_p, sizeof(pair_info), compare_pair_info );
/* print it */
fprintf(aliout, "%d sequence; length of alignment %d\n",
n_seq, (int) strlen(AS[0]));
fprintf(aliout, "alifold output\n");
for (k=0; pi[k].i>0; k++) {
pi[k].comp = (ptable[pi[k].i] == pi[k].j) ? 1:0;
print_pi(pi[k], aliout);
}
fprintf(aliout, "%s\n", mfe);
free(ptable);
free(pi);
}
PUBLIC int bp_distance(const char *str1, const char *str2)
{
/* dist = {number of base pairs in one structure but not in the other} */
/* same as edit distance with pair_open pair_close as move set */
int dist;
short i,l;
short *t1, *t2;
dist = 0;
t1 = make_pair_table(str1);
t2 = make_pair_table(str2);
l = (t1[0]<t2[0])?t1[0]:t2[0]; /* minimum of the two lengths */
for (i=1; i<=l; i++)
if (t1[i]!=t2[i]) {
if (t1[i]>i) dist++;
if (t2[i]>i) dist++;
}
free(t1); free(t2);
return dist;
}
PRIVATE char *annote(const char *structure, const char *AS[]) {
char *ps;
int i, n, s, maxl;
short *ptable;
make_pair_matrix();
n = strlen(AS[0]);
maxl = 1024;
ps = (char *) space(maxl);
ptable = make_pair_table(structure);
for (i=1; i<=n; i++) {
char pps[64], ci='\0', cj='\0';
int j, type, pfreq[8] = {0,0,0,0,0,0,0,0}, vi=0, vj=0;
if ((j=ptable[i])<i) continue;
for (s=0; AS[s]!=NULL; s++) {
type = pair[encode_char(AS[s][i-1])][encode_char(AS[s][j-1])];
pfreq[type]++;
if (type) {
if (AS[s][i-1] != ci) { ci = AS[s][i-1]; vi++;}
if (AS[s][j-1] != cj) { cj = AS[s][j-1]; vj++;}
}
}
if (maxl - strlen(ps) < 128) {
maxl *= 2;
ps = realloc(ps, maxl);
if (ps==NULL) nrerror("out of memory in realloc");
}
if (pfreq[0]>0) {
snprintf(pps, 64, "%d %d %d gmark\n", i, j, pfreq[0]);
strcat(ps, pps);
}
if (vi>1) {
snprintf(pps, 64, "%d cmark\n", i);
strcat(ps, pps);
}
if (vj>1) {
snprintf(pps, 64, "%d cmark\n", j);
strcat(ps, pps);
}
}
free(ptable);
return ps;
}
/*---------------------------------------------------------------------------*/
PUBLIC int xrna_plot(char *string, char *structure, char *ssfile)
{ /* produce input for XRNA RNA drawing program */
FILE *ss_file;
int i;
int length;
short *pair_table;
float *X, *Y;
ss_file = fopen(ssfile, "w");
if (ss_file == NULL) {
fprintf(stderr, "can't open file %s - not doing xy_plot\n", ssfile);
return 0;
}
length = strlen(string);
pair_table = make_pair_table(structure);
/* make coordinates */
X = (float *) space((length+1)*sizeof(float));
Y = (float *) space((length+1)*sizeof(float));
if (rna_plot_type == 0)
i = simple_xy_coordinates(pair_table, X, Y);
else
i = naview_xy_coordinates(pair_table, X, Y);
if (i!=length)
fprintf(stderr,"strange things happening in xrna_plot...\n");
fprintf(ss_file,
"# Vienna RNA Package %s, XRNA output\n"
"# CreationDate: %s\n"
"# Options: %s\n", VERSION, time_stamp(), option_string());
for (i=1; i<=length; i++)
/* XRNA likes to have coordinate mirrored, so we use (-X, Y) */
fprintf(ss_file, "%d %c %6.2f %6.2f %d %d\n", i, string[i-1],
-X[i-1], Y[i-1], (pair_table[i]?1:0), pair_table[i]);
fclose(ss_file);
free(pair_table);
free(X); free(Y);
return 1; /* success */
}
PRIVATE struct plist *b2plist(const char *struc) {
/* convert bracket string to plist */
short *pt;
struct plist *pl;
int i,k=0;
pt = make_pair_table(struc);
pl = (struct plist *)space(strlen(struc)/2*sizeof(struct plist));
for (i=1; i<strlen(struc); i++) {
if (pt[i]>i) {
pl[k].i = i;
pl[k].j = pt[i];
pl[k++].p = 0.95*0.95;
}
}
free(pt);
pl[k].i=0;
pl[k].j=0;
pl[k++].p=0.;
return pl;
}
PUBLIC int
browse_neighs(char *seq,
char *struc,
int verbosity_level,
int shifts,
int noLP,
int (*funct) (struct_en*, struct_en*)){
make_pair_matrix();
short int *s0 = encode_sequence(seq, 0);
short int *s1 = encode_sequence(seq, 1);
short int *str = make_pair_table(struc);
int res = browse_neighs_pt(seq, str, s0, s1, verbosity_level, shifts, noLP, funct);
free(s0);
free(s1);
free(str);
return res;
}
PUBLIC int gmlRNA(char *string, char *structure, char *ssfile, char option)
{
FILE *gmlfile;
int i;
int length;
int labels=0;
short *pair_table;
float *X, *Y;
if (isupper(option)) labels = 1;
gmlfile = fopen(ssfile, "w");
if (gmlfile == NULL) {
fprintf(stderr, "can't open file %s - not doing xy_plot\n", ssfile);
return 0;
}
length = strlen(string);
pair_table = make_pair_table(structure);
switch(option){
case 'X' :
case 'x' :
/* Simple XY Plot */
X = (float *) space((length+1)*sizeof(float));
Y = (float *) space((length+1)*sizeof(float));
if (rna_plot_type == 0)
i = simple_xy_coordinates(pair_table, X, Y);
else
i = naview_xy_coordinates(pair_table, X, Y);
if(i!=length) fprintf(stderr,"strange things happening in gmlRNA ...\n");
break;
default:
/* No Graphics Information */
X = NULL;
Y = NULL;
}
fprintf(gmlfile,
"# Vienna RNA Package %s\n"
"# GML Output\n"
"# CreationDate: %s\n"
"# Name: %s\n"
"# Options: %s\n", VERSION, time_stamp(), ssfile, option_string());
fprintf(gmlfile,
"graph [\n"
" directed 0\n");
for (i=1; i<=length; i++){
fprintf(gmlfile,
" node [ id %d ", i);
if (option) fprintf(gmlfile,
"label \"%c\"",string[i-1]);
if ((option == 'X')||(option=='x'))
fprintf(gmlfile,
"\n graphics [ x %9.4f y %9.4f ]\n", X[i-1], Y[i-1]);
fprintf(gmlfile," ]\n");
}
for (i=1; i<length; i++)
fprintf(gmlfile,
"edge [ source %d target %d ]\n", i, i+1);
for (i=1; i<=length; i++) {
if (pair_table[i]>i)
fprintf(gmlfile,
"edge [ source %d target %d ]\n", i, pair_table[i]);
}
fprintf(gmlfile, "]\n");
fclose(gmlfile);
free(pair_table);
free(X); free(Y);
return 1; /* success */
}
PUBLIC int ssv_rna_plot(char *string, char *structure, char *ssfile)
{ /* produce input for the SStructView java applet */
FILE *ssvfile;
int i, bp;
int length;
short *pair_table;
float *X, *Y;
float xmin, xmax, ymin, ymax;
ssvfile = fopen(ssfile, "w");
if (ssvfile == NULL) {
fprintf(stderr, "can't open file %s - not doing xy_plot\n", ssfile);
return 0;
}
length = strlen(string);
pair_table = make_pair_table(structure);
/* make coordinates */
X = (float *) space((length+1)*sizeof(float));
Y = (float *) space((length+1)*sizeof(float));
if (rna_plot_type == 0)
i = simple_xy_coordinates(pair_table, X, Y);
else
i = naview_xy_coordinates(pair_table, X, Y);
if (i!=length)
fprintf(stderr,"strange things happening in ssv_rna_plot...\n");
/* make coords nonegative */
xmin = xmax = X[0];
ymin = ymax = Y[0];
for (i = 1; i < length; i++) {
xmin = X[i] < xmin ? X[i] : xmin;
xmax = X[i] > xmax ? X[i] : xmax;
ymin = Y[i] < ymin ? Y[i] : ymin;
ymax = Y[i] > ymax ? Y[i] : ymax;
}
if (xmin<1) {
for (i = 0; i <= length; i++)
X[i] -= xmin-1;
xmin = 1;
}
if (ymin<1) {
for (i = 0; i <= length; i++)
Y[i] -= ymin-1;
ymin = 1;
}
#if 0
{
float size, xoff, yoff;
float JSIZE = 500; /* size of the java applet window */
/* rescale coordinates, center on square of size HSIZE */
size = MAX((xmax-xmin),(ymax-ymin));
xoff = (size - xmax + xmin)/2;
yoff = (size - ymax + ymin)/2;
for (i = 0; i <= length; i++) {
X[i] = (X[i]-xmin+xoff)*(JSIZE-10)/size + 5;
Y[i] = (Y[i]-ymin+yoff)*(JSIZE-10)/size + 5;
}
}
#endif
/* */
fprintf(ssvfile,
"# Vienna RNA Package %s\n"
"# SStructView Output\n"
"# CreationDate: %s\n"
"# Name: %s\n"
"# Options: %s\n", VERSION, time_stamp(), ssfile, option_string());
for (i=1; i<=length; i++)
fprintf(ssvfile, "BASE\t%d\t%c\t%d\t%d\n",
i, string[i-1], (int) (X[i-1]+0.5), (int) (Y[i-1]+0.5));
for (bp=1, i=1; i<=length; i++)
if (pair_table[i]>i)
fprintf(ssvfile, "BASE-PAIR\tbp%d\t%d\t%d\n", bp++, i, pair_table[i]);
fclose(ssvfile);
free(pair_table);
free(X); free(Y);
return 1; /* success */
}
int svg_rna_plot(char *string, char *structure, char *ssfile)
{
float xmin, xmax, ymin, ymax, size;
int i, length;
float *X, *Y;
FILE *xyplot;
short *pair_table;
length = strlen(string);
xyplot = fopen(ssfile, "w");
if (xyplot == NULL) {
fprintf(stderr, "can't open file %s - not doing xy_plot\n", ssfile);
return 0;
}
pair_table = make_pair_table(structure);
X = (float *) space((length+1)*sizeof(float));
Y = (float *) space((length+1)*sizeof(float));
if (rna_plot_type == 0)
i = simple_xy_coordinates(pair_table, X, Y);
else
i = naview_xy_coordinates(pair_table, X, Y);
if(i!=length) fprintf(stderr,"strange things happening in PS_rna_plot...\n");
xmin = xmax = X[0];
ymin = ymax = Y[0];
for (i = 1; i < length; i++) {
xmin = X[i] < xmin ? X[i] : xmin;
xmax = X[i] > xmax ? X[i] : xmax;
ymin = Y[i] < ymin ? Y[i] : ymin;
ymax = Y[i] > ymax ? Y[i] : ymax;
}
for (i = 0; i < length; i++)
Y[i] = ymin+ymax - Y[i]; /* mirror coordinates so they look as in PS */
size = MAX((xmax-xmin),(ymax-ymin));
size += 10; /* add some so the bounding box isn't too tight */
fprintf(xyplot,
"<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"yes\"?>\n"
"<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"452\" width=\"452\">\n");
fprintf(xyplot,
"<script type=\"text/ecmascript\">\n"
" <![CDATA[\n"
" var shown = 1;\n"
" function click() {\n"
" var seq = document.getElementById(\"seq\");\n"
" if (shown==1) {\n"
" seq.setAttribute(\"style\", \"visibility: hidden\");\n"
" shown = 0;\n"
" } else {\n"
" seq.setAttribute(\"style\", \"visibility: visible\");\n"
" shown = 1;\n"
" }\n"
" }\n"
" ]]>\n"
"</script>\n");
fprintf(xyplot,
" <rect style=\"stroke: white; fill: white\" height=\"452\" x=\"0\" y=\"0\" width=\"452\" onclick=\"click(evt)\" />\n"
" <g transform=\"scale(%7f,%7f) translate(%7f,%7f)\">\n",
SIZE/size, SIZE/size, (size-xmin-xmax)/2, (size-ymin-ymax)/2);
fprintf(xyplot,
" <polyline style=\"stroke: black; fill: none; stroke-width: 1.5\" id=\"outline\" points=\"\n");
for (i = 0; i < length; i++)
fprintf(xyplot, " %3.3f,%3.3f\n", X[i], Y[i]);
fprintf(xyplot," \" />\n");
fprintf(xyplot," <g style=\"stroke: black; stroke-width: 1\" id=\"pairs\">\n");
for (i = 1; i <= length; i++) {
int j;
if ((j=pair_table[i])>i)
fprintf(xyplot,
" <line id=\"%d,%d\" x1=\"%6.3f\" y1=\"%6.3f\" x2=\"%6.3f\" y2=\"%6.3f\" />\n",
i,j, X[i-1], Y[i-1], X[j-1], Y[j-1]);
}
fprintf(xyplot, " </g>\n");
fprintf(xyplot, " <g style=\"font-family: SansSerif\" transform=\"translate(-4.6, 4)\" id=\"seq\">\n");
for (i = 0; i < length; i++)
fprintf(xyplot, " <text x=\"%.3f\" y=\"%.3f\">%c</text>\n", X[i], Y[i], string[i]);
fprintf(xyplot, " </g>\n");
fprintf(xyplot, " </g>\n");
fprintf(xyplot, "</svg>\n");
fclose(xyplot);
free(pair_table);
free(X); free(Y);
return 1; /* success */
}
int PS_rna_plot_a(char *string, char *structure, char *ssfile, char *pre, char *post)
{
float xmin, xmax, ymin, ymax, size;
int i, length;
float *X, *Y;
FILE *xyplot;
short *pair_table;
char *c;
length = strlen(string);
xyplot = fopen(ssfile, "w");
if (xyplot == NULL) {
fprintf(stderr, "can't open file %s - not doing xy_plot\n", ssfile);
return 0;
}
pair_table = make_pair_table(structure);
X = (float *) space((length+1)*sizeof(float));
Y = (float *) space((length+1)*sizeof(float));
if (rna_plot_type == 0)
i = simple_xy_coordinates(pair_table, X, Y);
else
i = naview_xy_coordinates(pair_table, X, Y);
if(i!=length) fprintf(stderr,"strange things happening in PS_rna_plot...\n");
xmin = xmax = X[0];
ymin = ymax = Y[0];
for (i = 1; i < length; i++) {
xmin = X[i] < xmin ? X[i] : xmin;
xmax = X[i] > xmax ? X[i] : xmax;
ymin = Y[i] < ymin ? Y[i] : ymin;
ymax = Y[i] > ymax ? Y[i] : ymax;
}
size = MAX((xmax-xmin),(ymax-ymin));
fprintf(xyplot,
"%%!PS-Adobe-3.0 EPSF-3.0\n"
"%%%%Creator: %s, ViennaRNA-%s\n"
"%%%%CreationDate: %s"
"%%%%Title: RNA Secondary Structure Plot\n"
"%%%%BoundingBox: 66 210 518 662\n"
"%%%%DocumentFonts: Helvetica\n"
"%%%%Pages: 1\n"
"%%%%EndComments\n\n"
"%%Options: %s\n", rcsid+5, VERSION, time_stamp(), option_string());
fprintf(xyplot, "%% to switch off outline pairs of sequence comment or\n"
"%% delete the appropriate line near the end of the file\n\n");
fprintf(xyplot, "%s", RNAss_head);
if (pre || post) {
fprintf(xyplot, "%s", anote_macros);
}
fprintf(xyplot, "%%%%EndProlog\n");
fprintf(xyplot, "RNAplot begin\n"
"%% data start here\n");
if ((c = strchr(structure, '&'))) {
int cutpoint;
cutpoint = c - structure;
string[cutpoint] = ' '; /* replace & with space */
fprintf(xyplot, "/cutpoint %d def\n", cutpoint);
}
/* sequence */
fprintf(xyplot,"/sequence (\\\n");
i=0;
while (i<length) {
fprintf(xyplot, "%.255s\\\n", string+i); /* no lines longer than 255 */
i+=255;
}
fprintf(xyplot,") def\n");
/* coordinates */
fprintf(xyplot, "/coor [\n");
for (i = 0; i < length; i++)
fprintf(xyplot, "[%3.3f %3.3f]\n", X[i], Y[i]);
fprintf(xyplot, "] def\n");
/* base pairs */
fprintf(xyplot, "/pairs [\n");
for (i = 1; i <= length; i++)
if (pair_table[i]>i)
fprintf(xyplot, "[%d %d]\n", i, pair_table[i]);
fprintf(xyplot, "] def\n\n");
fprintf(xyplot, "init\n\n");
/* draw the data */
if (pre) {
fprintf(xyplot, "%% Start Annotations\n");
fprintf(xyplot, "%s\n", pre);
fprintf(xyplot, "%% End Annotations\n");
}
fprintf(xyplot,
"%% switch off outline pairs or bases by removing these lines\n"
"drawoutline\n"
"drawpairs\n"
"drawbases\n");
if (post) {
fprintf(xyplot, "%% Start Annotations\n");
fprintf(xyplot, "%s\n", post);
fprintf(xyplot, "%% End Annotations\n");
}
fprintf(xyplot, "%% show it\nshowpage\n");
fprintf(xyplot, "end\n");
fprintf(xyplot, "%%%%EOF\n");
fclose(xyplot);
free(pair_table);
free(X); free(Y);
return 1; /* success */
}
void DSU::ComputeTBD(TBD &pqueue, int maxkeep, int num_threshold, bool outer, bool noLP, bool shifts, bool debug, vector<RNAsaddle> *output_saddles, int conn_neighs, bool no_new)
{
int cnt = 0;
clock_t time_tbd = clock();
// go through all pairs in queue
while (pqueue.size()>0) {
// check time:
double time_secs = ((clock() - time)/(double)CLOCKS_PER_SEC);
if (stop_after && (time_secs > stop_after)) {
fprintf(stderr, "Time threshold reached (%d secs.), processed %d/%d\n", stop_after, cnt, pqueue.size()+cnt);
break;
}
// just visualisation
if (!output_saddles && cnt%100==0) {
double tim = (clock() - time_tbd)/(double)CLOCKS_PER_SEC;
std::pair<int, int> mem = getValue();
//double one = ((sizeof(char)*strlen(seq) + sizeof(short)*strlen(seq)) + sizeof(RNAsaddle)) / 1024.0;
fprintf(stderr, "Finding path: %7d/%7d; Time: %6.2f; Est.:%6.2f Mem.:%6.1fMB VM %6.1fMB PM\n", cnt, pqueue.size()+cnt, tim, tim/(double)cnt*pqueue.size(), mem.first/1024.0, mem.second/1024.0);
}
// apply threshold
if (cnt>num_threshold) {
fprintf(stderr, "Number threshold reached, processed %d/%d\n", cnt, pqueue.size()+cnt);
break;
} else {
cnt++;
}
// get next
TBDentry tbd = pqueue.get_first();
if (tbd.i==-1) {
fprintf(stderr, "Ending the path-finding -- i = %5d ; j = %5d ; fiber = %c ; type = %s \n", tbd.i, tbd.j, tbd.fiber?'Y':'N', type1_str[tbd.type_clust]);
break;
}
// check no-conn
if (conectivity.size() > 0 && !tbd.fiber && conectivity.joint(tbd.i, tbd.j)) continue;
// get path
if (debug) fprintf(stderr, "path between (%3d, %3d) type=%s fiber=%d:\n", tbd.i, tbd.j, type1_str[tbd.type_clust], tbd.fiber);
//2fprintf(stderr, "depth: %d\n%s\n%s\n%s\n", maxkeep, seq, LM[tbd.i].str_ch, LM[tbd.j].str_ch);
if (pknots) {
path_pk *path = get_path_light_pk(seq, LM[tbd.i].structure, LM[tbd.j].structure, maxkeep);
// variables for outer insertion
double max_energy= -1e8;
path_pk *max_path = path;
// variables for inner loops and insertions
// get the length of path for speed up
int length = 0;
for (path_pk *tmp = path; tmp && tmp->structure; tmp++) {
if (max_path->en < tmp->en) max_path = tmp;
length ++;
}
// create vector of known LM numbers on path (where 0 and length-1 are known)
vector<int> lm_numbers(length, -1);
lm_numbers[0] = tbd.i;
lm_numbers[length-1] = tbd.j;
// debug
if (debug) {
for (int i=0; i<length; i++) {
fprintf(stderr, "path[%3d] %s %6.2f\n", i, pt_to_str_pk(path[i].structure).c_str(), path[i].en/100.0);
}
}
// bisect the path and find new LMs:
unsigned int old_size = LM.size();
FindNumbers(0, length-1, path, lm_numbers, shifts, noLP, debug, no_new);
// if we have found new minima and we want to do more than simple reevaluation of path (--conn-neighs>0)
if (LM.size() - old_size > 0 && conn_neighs > 0 && !no_new) {
for (unsigned int j=old_size; j<LM.size(); j++) {
// sort 'em according to Hamming D. and take first "conn_neighs"
multimap<int, int> distances;
for (unsigned int i=0; i<old_size; i++) {
distances.insert(make_pair(HammingDist(LM[i].structure, LM[j].structure), i));
}
int cnt = 0;
int last_hd = -1;
for (auto it=distances.begin(); it!=distances.end(); it++) {
if (cnt > conn_neighs && last_hd != it->first) {
break;
}
pqueue.insert(it->second, j, EXPERIM, false);
cnt++;
last_hd = it->first;
}
}
}
// debug
if (debug) {
int diff = 1;
int last_num = lm_numbers[0];
for (int i=0; i<length; i++) {
fprintf(stderr, "path[%3d]= %4d (%s %6.2f)\n", i, lm_numbers[i], pt_to_str_pk(path[i].structure).c_str(), path[i].en/100.0);
if (lm_numbers[i]!=last_num && lm_numbers[i]!=-1) {
diff++;
last_num=lm_numbers[i];
}
}
histo[length][diff]++;
histo[length][0]++;
}
// now process the array of found numbers:
int last_num = lm_numbers[0];
for (int i=1; i<length; i++) {
if (lm_numbers[i]!=-1 && lm_numbers[i]!=last_num) {
// get the highest saddle in case we traveled through many "-1" saddles:
int j=i-1;
int highest_num = i;
while (j>0) {
// check if not higher saddle:
if (path[highest_num].en < path[j].en) highest_num = j;
// we found first that is not -1
if (lm_numbers[j]!=-1) break;
j--;
}
// save saddle
SDtype typ = (j==i-1?DIRECT:REDUCED);
RNAsaddle saddle(last_num, lm_numbers[i], typ);
saddle.energy = path[highest_num].en;
saddle.str_ch = NULL;
saddle.structure = allocopy(path[highest_num].structure);
bool inserted = InsertUB(saddle, debug);
// ???
if (output_saddles && inserted) {
output_saddles->push_back(saddle);
}
// try to insert new things into TBD:
if ((lm_numbers[i]!=lm_numbers[length-1] || lm_numbers[i-1]!=lm_numbers[0]) && !no_new) {
// check no-conn
if (conectivity.size() > 0) conectivity.union_set(tbd.i, tbd.j);
pqueue.insert(lm_numbers[i-1], lm_numbers[i], NEW_FOUND, true);
}
last_num = lm_numbers[i];
}
}
// insert saddle between outer structures
if (outer) {
RNAsaddle tmp(tbd.i, tbd.j, NOT_SURE);
tmp.energy = en_fltoi(max_energy);
tmp.str_ch = NULL;
tmp.structure = allocopy(max_path->structure);
bool inserted = InsertUB(tmp, debug);
if (output_saddles && inserted) {
output_saddles->push_back(tmp);
}
}
free_path_pk(path);
} else {
//fprintf(stderr, "%s\n%s\n%s\n", seq, LM[tbd.i].str_ch, LM[tbd.j].str_ch);
path_t *path = get_path(seq, LM[tbd.i].str_ch, LM[tbd.j].str_ch, maxkeep);
// variables for outer insertion
double max_energy= -1e8;
path_t *max_path = path;
// variables for inner loops and insertions
// get the length of path for speed up
int length = 0;
for (path_t *tmp = path; tmp && tmp->s; tmp++) {
if (max_path->en < tmp->en) max_path = tmp;
length ++;
}
// create vector of known LM numbers on path (where 0 and length-1 are known)
vector<int> lm_numbers(length, -1);
lm_numbers[0] = tbd.i;
lm_numbers[length-1] = tbd.j;
// bisect the path and find new LMs:
unsigned int old_size = LM.size();
FindNumbers(0, length-1, path, lm_numbers, shifts, noLP, debug, no_new);
// if we have found new minima and we want to do more than simple reevaluation of path (--conn-neighs>0)
if (LM.size() - old_size > 0 && conn_neighs > 0 && !no_new) {
for (unsigned int j=old_size; j<LM.size(); j++) {
// sort 'em according to Hamming D. and take first "conn_neighs"
multimap<int, int> distances;
for (unsigned int i=0; i<old_size; i++) {
distances.insert(make_pair(HammingDist(LM[i].structure, LM[j].structure), i));
}
int cnt = 0;
int last_hd = -1;
for (auto it=distances.begin(); it!=distances.end(); it++) {
if (cnt > conn_neighs && last_hd != it->first) {
break;
}
pqueue.insert(it->second, j, EXPERIM, false);
cnt++;
last_hd = it->first;
}
}
}
// debug
if (debug) {
int diff = 1;
int last_num = lm_numbers[0];
for (int i=0; i<length; i++) {
fprintf(stderr, "path[%3d]= %4d (%s %6.2f)\n", i, lm_numbers[i], path[i].s, path[i].en);
if (lm_numbers[i]!=last_num && lm_numbers[i]!=-1) {
diff++;
last_num=lm_numbers[i];
}
}
histo[length][diff]++;
histo[length][0]++;
}
// now process the array of found numbers:
int last_num = lm_numbers[0];
for (int i=1; i<length; i++) {
if (lm_numbers[i]!=-1 && lm_numbers[i]!=last_num) {
// get the highest saddle in case we traveled through many "-1" saddles:
int j=i-1;
int highest_num = i;
while (j>0) {
// check if not higher saddle:
if (path[highest_num].en < path[j].en) highest_num = j;
// we found first that is not -1
if (lm_numbers[j]!=-1) break;
j--;
}
// save saddle
SDtype typ = (j==i-1?DIRECT:REDUCED);
RNAsaddle saddle(last_num, lm_numbers[i], typ);
saddle.energy = en_fltoi(path[highest_num].en);
saddle.str_ch = NULL;
saddle.structure = make_pair_table(path[highest_num].s);
bool inserted = InsertUB(saddle, debug);
// ???
if (output_saddles && inserted) {
output_saddles->push_back(saddle);
}
// try to insert new things into TBD:
if ((lm_numbers[i]!=lm_numbers[length-1] || lm_numbers[i-1]!=lm_numbers[0]) && !no_new) {
// check no-conn
if (conectivity.size() > 0) conectivity.union_set(tbd.i, tbd.j);
pqueue.insert(lm_numbers[i-1], lm_numbers[i], NEW_FOUND, true);
}
last_num = lm_numbers[i];
}
}
// insert saddle between outer structures
if (outer) {
RNAsaddle tmp(tbd.i, tbd.j, NOT_SURE);
tmp.energy = en_fltoi(max_energy);
tmp.str_ch = NULL;
tmp.structure = make_pair_table(max_path->s);
bool inserted = InsertUB(tmp, debug);
if (output_saddles && inserted) {
output_saddles->push_back(tmp);
}
}
free_path(path);
}
// free stuff
//if (last_str) free(last_str);
} // all doing while
fprintf(stderr, "The end of finding paths(%d). Size of pqueue = %d\n", cnt, (int)pqueue.size());
}
void RNAFuncs::generateRNAAlignmentXML(const string &structure, const string &altStructure, const string &seq1, const string &seq2, const string &strname1, const string &strname2, ostream &s)
{
string base;
Uint i;
float *X, *Y;
short *pair_table,*alt_pair_table;
Uint basenr_x=1, basenr_y=1;
string filename;
X = new float[structure.size()];
Y = new float[structure.size()];
assert(seq1.size() == seq2.size());
assert(seq1.size() == structure.size());
assert(seq1.size() == altStructure.size());
// calculate coordinates
pair_table = make_pair_table(structure.c_str());
alt_pair_table = make_pair_table(altStructure.c_str());
i = naview_xy_coordinates(pair_table, X, Y);
if(i!=structure.size())
cerr << "strange things happening in squigglePlot ..." << endl;
// generate XML
s << "<alignment xname=\"" << strname1 << "\" yname=\"" << strname2 << "\">" << endl;
// seqalignment
s << " <seqalignment>" << endl;
for(i=0;i<seq1.length();i++)
{
s << " <base id=\"" << i+1 << "\" xbasenr=\"" << basenr_x << "\" ybasenr=\"" << basenr_y << "\" xbase=\"" << seq1[i] <<"\" ybase=\"" << seq2[i] << "\" />" << endl;
if(seq1[i]!='-')
basenr_x++;
if(seq2[i]!='-')
basenr_y++;
}
s << " </seqalignment>" << endl;
// pairs
s << " <pairs>" << endl;
for(i=0;i<structure.size();i++)
{
// both
if((unsigned short)pair_table[i+1]>i+1 && (unsigned short)alt_pair_table[i+1]>i+1)
{
s << " <pair drawbaseid1=\"" << i+1 << "\" drawbaseid2=\"" << pair_table[i+1] << "\"/>" << endl;
}
else
{
if((unsigned short)pair_table[i+1]>i+1)
{
s << " <pair drawbaseid1=\"" << i+1 << "\" drawbaseid2=\"" << pair_table[i+1] << "\" structid=\"1\"/>" << endl;
}
else
{
if((unsigned short)alt_pair_table[i+1]>i+1)
s << " <pair drawbaseid1=\"" << i+1 << "\" drawbaseid2=\"" << alt_pair_table[i+1] << "\" structid=\"2\"/>" << endl;
}
}
}
s << " </pairs>" << endl;
// coordinates
// x structure
s << " <structure id=\"1\">" << endl;
s << " <drawbases>" << endl;
for(i=0;i<seq1.length();i++)
{
s << " <drawbase id=\"" << i+1 << "\" xcoor=\"" << X[i] << "\" ycoor=\"" << -Y[i] << "\"/>" << endl;
}
s << " </drawbases>" << endl;
s << " </structure>" << endl;
DELETE(X);
DELETE(Y);
X = new float[altStructure.size()];
Y = new float[altStructure.size()];
// y structure
i = naview_xy_coordinates(alt_pair_table, X, Y);
if(i!=structure.size())
cerr << "strange things happening in squigglePlot ..." << endl;
s << " <structure id=\"2\">" << endl;
s << " <drawbases>" << endl;
for(i=0;i<seq1.length();i++)
{
s << " <drawbase id=\"" << i+1 << "\" xcoor=\"" << X[i] << "\" ycoor=\"" << -Y[i] << "\"/>" << endl;
}
s << " </drawbases>" << endl;
s << " </structure>" << endl;
s << "</alignment>" << endl;
free(pair_table);
free(alt_pair_table);
DELETE(X);
DELETE(Y);
}
void RNAProfileAlignment::squigglePlot(const string &filename, SquigglePlotOptions &options) const
{
const double base_fontsize=8;
const Uint num_grey_colors=100;
const double min_grey_color=1.0;
string seq,structure;
string base,structname;
float *X,*Y,min_X=0,max_X=0,min_Y=0,max_Y=0;
Uint i;
short *pair_table;
int id_PS,id;
int ps_grey_colors[num_grey_colors];
int ps_color_red;
int ps_color_black;
double xpos,ypos;
deque<double> pairprob;
deque<BaseProbs> baseprobs;
getStructureAlignment(options.minPairProb,structure,pairprob);
getSequenceAlignment(baseprobs);
// filterConsensus(structure,pairprob,baseprobs,0.5);
//assert(baseprobs.size() == structure.size());
if(baseprobs.size() != structure.size())
cerr << "Error in resolving consensus structure!" << endl;
X = new float[structure.size()];
Y = new float[structure.size()];
pair_table = make_pair_table(structure.c_str());
i = naview_xy_coordinates(pair_table, X, Y);
if(i!=structure.size())
cerr << "strange things happening in squigglePlot ..." << endl;
// calculate image dimesions
for(i=0;i<structure.size();i++)
{
min_X=min(min_X,X[i]);
max_X=max(max_X,X[i]);
min_Y=min(min_Y,Y[i]);
max_Y=max(max_Y,Y[i]);
}
// id_PS = g2_open_PS("ali.ps", g2_A4, g2_PS_port);
id_PS = g2_open_EPSF((char*)filename.c_str());
id = g2_open_vd();
g2_attach(id,id_PS);
// cout << "min_X: " << min_X <<",max_X: " << max_X << ",min_Y: " << min_Y << "max_Y: " << max_Y << endl;
g2_set_coordinate_system(id_PS,595/2.0,842/2.0,0.5,0.5);
g2_set_line_width(id,0.2);
// set colors
double intv=min_grey_color/(double)num_grey_colors;
for(i=0;i<num_grey_colors;i++)
{
double grey_color=min_grey_color-i*intv;
ps_grey_colors[i]=g2_ink(id_PS,grey_color,grey_color,grey_color);
}
ps_color_black=g2_ink(id_PS,0,0,0);
if(options.greyColors)
ps_color_red=g2_ink(id_PS,0,0,0);
else
ps_color_red=g2_ink(id_PS,1,0,0);
// draw sequence
g2_set_font_size(id,base_fontsize);
for(i=0;i<structure.size();i++)
{
if(options.mostLikelySequence)
{
double p=getMlBaseFreq(baseprobs[i]);
//base color
if(p==1)
g2_pen(id,ps_color_red);
else
g2_pen(id,ps_grey_colors[(int)floor(p*num_grey_colors-1)]);
base=getMlBase(baseprobs[i]);
xpos=X[i]-base.length()*base_fontsize/2.0;
ypos=Y[i]-4;
g2_string(id,xpos,ypos,(char*)base.c_str());
}
else
{
drawBaseCircles(id_PS,baseprobs[i],X[i],Y[i]);
}
// connection to next base
if(i<structure.size()-1)
{
if((1-baseprobs[i].gap)*(1-baseprobs[i+1].gap)==1)
g2_pen(id,ps_color_red);
else
g2_pen(id,ps_grey_colors[(int)floor((1-baseprobs[i].gap)*(1-baseprobs[i+1].gap)*num_grey_colors-1)]);
g2_line(id,X[i],Y[i],X[i+1],Y[i+1]);
}
}
// draw pairings
// !!! pair_table indexing begins at 1 !!!
for(i=0;i<structure.size();i++)
{
if((unsigned short)pair_table[i+1]>i+1)
{
// pairs in both structures
if(pairprob[i]==1)
g2_pen(id,ps_color_red);
else
g2_pen(id,ps_grey_colors[(int)floor(pairprob[i]*num_grey_colors-1)]);
g2_line(id,X[i],Y[i],X[pair_table[i+1]-1],Y[pair_table[i+1]-1]);
}
}
g2_flush(id);
g2_close(id);
free(pair_table);
DELETE(X);
DELETE(Y);
}
int main(int arfc, char **arfv)
{
/*
arfv[1] is the input file with sequence on the first line and constraint on the second line
takes constraints of the ( ) and [ ] variety
arfv[2] is the list of structures output by subopt to check
*/
if (arfc != 3) {
usage(arfv[0]);
exit(1);
}
FILE *fin, *fout;
int i, *pos, *neg;
char *C;
if (arfv[1][0] == '-' && arfv[1][1] == '\0')
fin = stdin;
else if (!(fin = fopen(arfv[1], "rb"))) {
fprintf(stderr, "unable to open %s, bailing\n", arfv[1]);
exit(1);
}
if (arfv[2][0] == '-' && arfv[2][1] == '\0')
fout = stdin;
else if (!(fout = fopen(arfv[2], "rb"))) {
fprintf(stderr, "unable to open %s, bailing\n", arfv[2]);
if (stdin != fin)
fclose(fin);
exit(1);
}
skip_sequence(fin);
C = get_structure(fin);
make_pair_table(C, &pos, &neg);
printf("constraint string:\n%s\n", C);
free(C);
printf("forced pairs:");
for (i = 1; i <= len; ++i) {
if (pos[i])
printf(" (%d,%d)", i, pos[i]);
}
printf("\n");
printf("forbidden pairs:");
for (i = 1; i <= len; ++i) {
if (neg[i])
printf(" (%d,%d)", i, neg[i]);
}
printf("\n");
while (!feof(fout)) {
char *s;
int *ps;
s = get_structure(fout);
make_pair_table(s, &ps, NULL);
if (s[0]) {
compare(s, ps, pos, neg);
}
free(s);
free(ps);
}
free(pos);
free(neg);
if (stdin != fin)
fclose(fin);
if (stdin != fin)
fclose(fout);
return 0;
}
int main(int argc, char *argv[])
{
struct RRNAfold_args_info args_info;
/* let's call our cmdline parser */
if (RRNAfold_parser (argc, argv, &args_info) != 0)
exit(1) ;
// for ( unsigned i = 0 ; i < args_info.inputs_num ; ++i )
// printf("%s\n",args_info.inputs[i]) ;
int type = args_info.fold_type_arg;
char *iname = args_info.input_file_arg;
char *oname = args_info.output_file_arg;
double temp = args_info.temperature_arg;
printf("Using temp: %f",temp);
int length,n,i,j;
/* Reads in a file containing(-I) one sequence per line and outputs (-O) a file containing the corresponiding structures
*/
char *let,*seq1, *struct1,* struct2,* xstruc,*ffname;
float *x,*y,e1, e2, tree_dist, string_dist, profile_dist, kT,**pf1, **pf2;
FILE *fi,*fo;
short *pair_table;
/* open the file */
fi = fopen(iname, "r");
fo = fopen(oname, "a");
if (fi == NULL) {printf("I couldn't open results.dat for reading.\n");}
i=0;//Index used to keep track of line in the file
/* Scanning each line in the file and folding the given sequence */
seq1 = (char* ) space(sizeof(char)*(1000+1));
while (fscanf(fi, "%s\n",seq1) == 1){
printf("%s\n",seq1);
length=strlen(seq1);
/* fold at 30C instead of the default 37C */
temperature = temp; /* must be set *before* initializing */
update_fold_params();
/* allocate memory for fold(), could be skipped */
//initialize_fold(strlen(seq1));
/* allocate memory for structure and fold */
struct1 = (char* ) space(sizeof(char)*(length+1));
let = (char* ) space(sizeof(char)*(length+1));
x = (float* ) space(sizeof(float)*(length+1));
y = (float* ) space(sizeof(float)*(length+1));
/* Folding the sequence */
e1 = fold(seq1, struct1);
// fprintf(fo,"%s\n",struct1);
pair_table=make_pair_table(struct1);
if(type==2){
j=simple_xy_coordinates(pair_table,x,y);
}else{
j=naview_xy_coordinates(pair_table,x,y);
}
j=0;
let = seq1;
while(j < (length)){
fprintf(fo,"%d,%f,%f,%c,%d,%d\n",i,x[j],y[j],*let,j,(pair_table[(j+1)]-1));
let++;
j++;
}
/* Makes .ps file of RNA fold */
///////////////////////////////////////////
/*
ffname=(char*) space(sizeof(char)*10);
n=sprintf(ffname, "rna%d.ps",i);
printf("%s\n",ffname);
(void) PS_rna_plot(seq1, struct1, ffname);
*/
///////////////////////////////////////////
free_arrays(); /* free arrays used in fold() */
i++;
} /* close the file */
fclose(fi);
fclose(fo);
}
/* len1 is the length of the LONGER input seq ! */
void seperate_bp(char **inter, int len1, char **intra_l, char **intra_s) {
int i,j,len;
short *pt=NULL;
char *temp_inter, *pt_inter;
len=strlen((*inter));
/* printf("inter\n%s\n",(*inter)); */
i = len+1;
temp_inter=(char*)space(sizeof(char)*i);
/* to make a pair_table convert <|> to (|) */
pt_inter=(char*)space(sizeof(char)*i);
/* if shorter seq is first seq in constrained string, write the
longer one as the first one */
temp_inter[strlen((*inter))] = '\0';
pt_inter[strlen((*inter))] = '\0';
if (cut_point < len1) {
/* write the constrain for the longer seq first */
for (j=0,i=cut_point-1;i<len;i++,j++) {
switch ((*inter)[i]){
case '(':
temp_inter[j] = ')';
pt_inter[j] = ')';
break;
case ')':
temp_inter[j] = '(';
pt_inter[j] = '(';
break;
default:
temp_inter[j] = (*inter)[i];
pt_inter[j] = '.';
}
}
/* then add the constrain for the shorter seq */
for (i=0;i< cut_point-1;i++,j++) {
switch ((*inter)[i]){
case '(':
temp_inter[j] = ')';
pt_inter[j] = ')';
break;
case ')':
temp_inter[j] = '(';
pt_inter[j] = '(';
break;
default:
temp_inter[j] = (*inter)[i];
pt_inter[j] = '.';
}
}
cut_point = len1+1;
strcpy((*inter),temp_inter);
} else {
for (i=0;i<strlen((*inter));i++) {
switch ((*inter)[i]){
case '(':
pt_inter[i] = '(';
break;
case ')':
pt_inter[i] = ')';
break;
default:
pt_inter[i] = '.';
}
}
}
pt = make_pair_table(pt_inter);
/* intramolecular structure in longer (_l) and shorter (_s) seq */
(*intra_l)=(char*)space(sizeof(char)*(len1+1));
(*intra_s)=(char*)space(sizeof(char)*(strlen((*inter))-len1+2));
(*intra_l)[len1] = '\0';
(*intra_s)[strlen((*inter))-len1+1] = '\0';
/* now seperate intermolecular from intramolecular bp */
for (i=1;i<=pt[0];i++) {
if (pt[i] == 0) {
temp_inter[i-1] = (*inter)[i-1];
if (i<cut_point) {
(*intra_l)[i-1] = (*inter)[i-1];
if ((*inter)[i-1] == '|')
(*intra_l)[i-1] = '.';
} else {
(*intra_s)[i-cut_point] = (*inter)[i-1];
if ((*inter)[i-1] == '|')
(*intra_s)[i-cut_point] = '.';
}
} else {
if (i<cut_point) {
/* intermolekular bp */
if (pt[i]>=cut_point){
temp_inter[i-1] = (*inter)[i-1];
(*intra_l)[i-1] = '.';
(*intra_s)[pt[i]-cut_point] = '.';
} else { /* intramolekular bp */
(*intra_l)[i-1] = (*inter)[i-1];
temp_inter[i-1] = '.';
}
} else { /* i>=cut_point */
/* intermolekular bp */
if (pt[i] < cut_point){
temp_inter[i-1] = (*inter)[i-1];
/* (*intra_s)[i-1] = '.'; */
} else { /* intramolekular bp */
(*intra_s)[i-cut_point] = (*inter)[i-1];
temp_inter[i-1] = '.';
}
}
}
}
/* printf("%s -1\n%s -2\n%s -3\n%s -4\n",(*inter),temp_inter,(*intra_l),(*intra_s)); */
strcpy((*inter),temp_inter);
free(temp_inter);
free(pt_inter);
free(pt);
}
void encode_str(encoded *enc, const char *str)
{
if (enc->pt) free(enc->pt);
enc->pt = make_pair_table(str);
}
void RNAFuncs::drawRNAStructure(const string &seq, const string &structure, const string &filename_prefix, const string &structname, const list<pair<Uint,Uint> > ®ions, const SquigglePlotOptions &options)
{
const double base_fontsize=12;
float *X, *Y,min_X=0,max_X=0,min_Y=0,max_Y=0;
Uint i;
short *pair_table;
int id_PS,id_FIG=0,id;
int color_black, *colors;
double xpos,ypos;
char buf[10];
string filename;
double *points;
int numPoints;
#ifdef HAVE_LIBGD
int id_PNG=0,id_JPG=0;
#endif
X = new float[structure.size()];
Y = new float[structure.size()];
points=new double[2*structure.size()];
colors=new int[NUM_COLORS];
assert(seq.size() == structure.size());
pair_table = make_pair_table(structure.c_str());
i = naview_xy_coordinates(pair_table, X, Y);
if(i!=structure.size())
cerr << "strange things happening in squigglePlot ..." << endl;
// scale image
for(i=0;i<structure.size();i++)
{
X[i]*=static_cast<float>(options.scale);
Y[i]*=static_cast<float>(options.scale);
}
// calculate image dimensions
for(i=0;i<structure.size();i++)
{
min_X=min(min_X,X[i]);
max_X=max(max_X,X[i]);
min_Y=min(min_Y,Y[i]);
max_Y=max(max_Y,Y[i]);
}
// add a border to image size
min_X-=10;
max_X+=10;
min_Y-=10;
max_Y+=10;
//id_PS = g2_open_PS("ali.ps", g2_A4, g2_PS_port);
filename=filename_prefix + ".ps";
id_PS = g2_open_EPSF((char*)filename.c_str());
g2_set_coordinate_system(id_PS,-min_X,-min_Y,1,1);
if(options.generateFIG)
{
filename=filename_prefix + ".fig";
id_FIG=g2_open_FIG((char*)filename.c_str());
g2_set_coordinate_system(id_FIG,-min_X,-min_Y,1,1);
}
#ifdef HAVE_LIBGD
if(options.generatePNG)
{
filename=filename_prefix + ".png";
id_PNG=g2_open_gd((char*)filename.c_str(),(int)(max_X-min_X),(int)(max_Y-min_Y),g2_gd_png);
g2_set_coordinate_system(id_PNG,-min_X,-min_Y,1,1);
}
if(options.generateJPG)
{
filename=filename_prefix + ".jpg";
id_JPG=g2_open_gd((char*)filename.c_str(),(int)(max_X-min_X),(int)(max_Y-min_Y),g2_gd_jpeg);
g2_set_coordinate_system(id_PS,-min_X,-min_Y,1,1);
}
#endif
id = g2_open_vd();
g2_attach(id,id_PS);
if(options.generateFIG)
{
g2_attach(id,id_FIG);
}
#ifdef HAVE_LIBGD
if(options.generatePNG)
g2_attach(id,id_PNG);
if(options.generateJPG)
g2_attach(id,id_JPG);
#endif
// cout << "min_X: " << min_X <<",max_X: " << max_X << ",min_Y: " << min_Y << "max_Y: " << max_Y << endl;
// g2_set_coordinate_system(id_PS,595/2.0,842/2.0,0.5,0.5);
// define colors
if(options.greyColors)
{
color_black=g2_ink(id_PS,0,0,0);
for(i=0;i<NUM_COLORS;i++)
colors[i]=g2_ink(id_PS,0,0,0);
if(options.generateFIG)
{
color_black=g2_ink(id_FIG,0,0,0);
for(i=0;i<NUM_COLORS;i++)
colors[i]=g2_ink(id_FIG,0,0,0);
}
#ifdef HAVE_LIBGD
if(options.generatePNG)
{
color_black=g2_ink(id_PNG,0,0,0);
for(i=0;i<NUM_COLORS;i++)
colors[i]=g2_ink(id_PNG,0,0,0);
}
if(options.generateJPG)
{
color_black=g2_ink(id_JPG,0,0,0);
for(i=0;i<NUM_COLORS;i++)
colors[i]=g2_ink(id_JPG,0,0,0);
}
#endif
}
else
{
color_black=g2_ink(id,0,0,0);
colors[COLOR_RED]=g2_ink(id_PS,COLOR_DEF_RED);
colors[COLOR_GREEN]=g2_ink(id_PS,COLOR_DEF_GREEN);
colors[COLOR_BLUE]=g2_ink(id_PS,COLOR_DEF_BLUE);
colors[COLOR_YELLOW]=g2_ink(id_PS,COLOR_DEF_YELLOW);
colors[COLOR_MAGENTA]=g2_ink(id_PS,COLOR_DEF_MAGENTA);
colors[COLOR_TURKIS]=g2_ink(id_PS,COLOR_DEF_TURKIS);
if(options.generateFIG)
{
color_black=g2_ink(id_FIG,0,0,0);
colors[COLOR_RED]=g2_ink(id_FIG,COLOR_DEF_RED);
colors[COLOR_GREEN]=g2_ink(id_FIG,COLOR_DEF_GREEN);
colors[COLOR_BLUE]=g2_ink(id_FIG,COLOR_DEF_BLUE);
colors[COLOR_YELLOW]=g2_ink(id_FIG,COLOR_DEF_YELLOW);
colors[COLOR_MAGENTA]=g2_ink(id_FIG,COLOR_DEF_MAGENTA);
colors[COLOR_TURKIS]=g2_ink(id_FIG,COLOR_DEF_TURKIS);
}
#ifdef HAVE_LIBGD
if(options.generatePNG)
{
color_black=g2_ink(id_PNG,0,0,0);
colors[COLOR_RED]=g2_ink(id_PNG,COLOR_DEF_RED);
colors[COLOR_GREEN]=g2_ink(id_PNG,COLOR_DEF_GREEN);
colors[COLOR_BLUE]=g2_ink(id_PNG,COLOR_DEF_BLUE);
colors[COLOR_YELLOW]=g2_ink(id_PNG,COLOR_DEF_YELLOW);
colors[COLOR_MAGENTA]=g2_ink(id_PNG,COLOR_DEF_MAGENTA);
colors[COLOR_TURKIS]=g2_ink(id_PNG,COLOR_DEF_TURKIS);
}
if(options.generateJPG)
{
color_black=g2_ink(id_JPG,0,0,0);
colors[COLOR_RED]=g2_ink(id_JPG,COLOR_DEF_RED);
colors[COLOR_GREEN]=g2_ink(id_JPG,COLOR_DEF_GREEN);
colors[COLOR_BLUE]=g2_ink(id_JPG,COLOR_DEF_BLUE);
colors[COLOR_YELLOW]=g2_ink(id_JPG,COLOR_DEF_YELLOW);
colors[COLOR_MAGENTA]=g2_ink(id_JPG,COLOR_DEF_MAGENTA);
colors[COLOR_TURKIS]=g2_ink(id_JPG,COLOR_DEF_TURKIS);
}
#endif
}
for(i=0;i<structure.size();i++)
{
// connection to next base
if(i<structure.size()-1)
{
// if the connected bases are only in one of the structures
// use the appropriate color
g2_line(id,X[i],Y[i],X[i+1],Y[i+1]);
// draw circles at line endpoints
g2_filled_circle(id,X[i],Y[i],0.7*options.scale); // circles are drawn twice, but thats ok ...
}
}
// draw pairings
// !!! pair_table indexing begins at 1 !!!
for(i=0;i<structure.size();i++)
{
if((unsigned short)pair_table[i+1]>i+1)
{
// pairs in both structures
g2_line(id,X[i],Y[i],X[pair_table[i+1]-1],Y[pair_table[i+1]-1]);
}
}
// draw regions
g2_set_line_width(id,0.4);
list<pair<Uint,Uint> >::const_iterator it;
Uint regionNr=0;
for(it=regions.begin();it!=regions.end();it++)
{
double center_x=0,center_y=0;
// fill coordinate list
for(i=0;i < it->second;i++)
{
points[2*i]=X[it->first+i];
points[2*i+1]=Y[it->first+i];
center_x+=X[it->first+i];
center_y+=Y[it->first+i];
}
numPoints=it->second-1;
center_x/=numPoints; // center of gravity
center_y/=numPoints;
g2_pen(id,colors[regionNr % NUM_COLORS]);
g2_poly_line(id,numPoints,points);
sprintf(buf,"%d",regionNr+1);
g2_string(id,center_x,center_y,buf); // draw region number
regionNr++;
}
// mark 5' end
g2_pen(id,color_black);
g2_string(id,X[0]-20,Y[0],"5'");
g2_set_font_size(id,base_fontsize*options.scale);
g2_set_line_width(id,0.2);
// draw sequence
for(i=0;i<structure.size();i++)
{
g2_pen(id,color_black); // match
xpos=X[i]-(base_fontsize*options.scale)/2;
ypos=Y[i]-4;
sprintf(buf,"%c",seq[i]);
g2_string(id,xpos,ypos,buf);
/*
if(!options.hideBaseNumbers)
{
// draw base number
if(basenr % options.baseNumInterval == 0)
{
sprintf(buf,"%d",basenr);
g2_string(id,xpos-20,ypos,buf);
}
}*/
}
// draw structure name
g2_string(id,min_X,max_Y-10,(char*)structname.c_str());
g2_flush(id);
g2_close(id);
free(pair_table);
DELETE(X);
DELETE(Y);
DELETE(points);
DELETE(colors);
}
int main(int argc, char *argv[])
{
char *string;
char *structure=NULL;
char *cstruc=NULL;
char *ns_bases=NULL;
char *c;
int n_seq;
int i;
int length;
int sym;
int endgaps = 0;
int mis = 0;
double min_en;
double real_en;
double sfact = 1.07;
int pf = 0;
int istty;
char *AS[MAX_NUM_NAMES]; /* aligned sequences */
char *names[MAX_NUM_NAMES]; /* sequence names */
AjPSeqset seq = NULL;
AjPFile confile = NULL;
AjPFile alifile = NULL;
AjPFile paramfile = NULL;
AjPFile outf = NULL;
AjPFile essfile = NULL;
AjPFile dotfile = NULL;
AjPStr constring = NULL;
float eT = 0.;
AjBool eGU;
AjBool eclose;
AjBool lonely;
AjPStr ensbases = NULL;
AjBool etloop;
AjPStr eenergy = NULL;
char ewt = '\0';
float escale = 0.;
AjPStr edangles = NULL;
char edangle = '\0';
ajint len;
AjPSeq tseq = NULL;
AjPStr tname = NULL;
int circ = 0;
int doAlnPS = 0;
int doColor = 0;
embInitPV("vrnaalifoldpf",argc,argv,"VIENNA",VERSION);
constring = ajStrNew();
seq = ajAcdGetSeqset("sequence");
confile = ajAcdGetInfile("constraintfile");
paramfile = ajAcdGetInfile("paramfile");
eT = ajAcdGetFloat("temperature");
eGU = ajAcdGetBoolean("gu");
eclose = ajAcdGetBoolean("closegu");
lonely = ajAcdGetBoolean("lp");
ensbases = ajAcdGetString("nsbases");
etloop = ajAcdGetBoolean("tetraloop");
eenergy = ajAcdGetListSingle("energy");
escale = ajAcdGetFloat("scale");
edangles = ajAcdGetListSingle("dangles");
mis = !!ajAcdGetBoolean("most");
endgaps = !!ajAcdGetBoolean("endgaps");
nc_fact = (double) ajAcdGetFloat("nspenalty");
cv_fact = (double) ajAcdGetFloat("covariance");
outf = ajAcdGetOutfile("outfile");
essfile = ajAcdGetOutfile("ssoutfile");
alifile = ajAcdGetOutfile("alignoutfile");
circ = !!ajAcdGetBoolean("circular");
doColor = !!ajAcdGetBoolean("colour");
dotfile = ajAcdGetOutfile("dotoutfile");
do_backtrack = 1;
pf = 1;
string = NULL;
istty = 0;
dangles = 2;
temperature = (double) eT;
noGU = (eGU) ? 0 : 1;
no_closingGU = (eclose) ? 0 : 1;
noLonelyPairs = (lonely) ? 0 : 1;
ns_bases = (ajStrGetLen(ensbases)) ? MAJSTRGETPTR(ensbases) : NULL;
tetra_loop = !!etloop;
ewt = *ajStrGetPtr(eenergy);
if(ewt == '0')
energy_set = 0;
else if(ewt == '1')
energy_set = 1;
else if(ewt == '2')
energy_set = 2;
sfact = (double) escale;
edangle = *ajStrGetPtr(edangles);
if(edangle == '0')
dangles = 0;
else if(edangle == '1')
dangles = 1;
else if(edangle == '2')
dangles = 2;
else if(edangle == '3')
dangles = 3;
if(paramfile)
read_parameter_file(paramfile);
if (ns_bases != NULL)
{
nonstandards = space(33);
c=ns_bases;
i=sym=0;
if (*c=='-')
{
sym=1;
c++;
}
while (*c!='\0')
{
if (*c!=',')
{
nonstandards[i++]=*c++;
nonstandards[i++]=*c;
if ((sym)&&(*c!=*(c-1)))
{
nonstandards[i++]=*c;
nonstandards[i++]=*(c-1);
}
}
c++;
}
}
if(alifile)
doAlnPS = 1;
if(confile)
vienna_GetConstraints(confile,&constring);
n_seq = ajSeqsetGetSize(seq);
if(n_seq > MAX_NUM_NAMES - 1)
ajFatal("[e]RNAalifold is restricted to %d sequences\n",
MAX_NUM_NAMES - 1);
if (n_seq==0)
ajFatal("No sequences found");
for(i=0;i<n_seq;++i)
{
tseq = (AjPSeq) ajSeqsetGetseqSeq(seq,i);
ajSeqGapStandard(tseq, '-');
tname = (AjPStr) ajSeqsetGetseqNameS(seq,i);
len = ajSeqGetLen(tseq);
AS[i] = (char *) space(len+1);
names[i] = (char *) space(ajStrGetLen(tname)+1);
strcpy(AS[i],ajSeqGetSeqC(tseq));
strcpy(names[i],ajStrGetPtr(tname));
}
AS[n_seq] = NULL;
names[n_seq] = NULL;
if (endgaps)
for (i=0; i<n_seq; i++)
mark_endgaps(AS[i], '~');
length = (int) strlen(AS[0]);
structure = (char *) space((unsigned) length+1);
if(confile)
{
fold_constrained = 1;
strcpy(structure,ajStrGetPtr(constring));
}
if (circ && noLonelyPairs)
ajWarn(
"warning, depending on the origin of the circular sequence, "
"some structures may be missed when using -noLP\n"
"Try rotating your sequence a few times\n");
if (circ)
min_en = circalifold((const char **)AS, structure);
else
min_en = alifold(AS, structure);
{
int i;
double s=0;
extern int eos_debug;
eos_debug=-1; /* shut off warnings about nonstandard pairs */
for (i=0; AS[i]!=NULL; i++)
if (circ)
s += energy_of_circ_struct(AS[i], structure);
else
s += energy_of_struct(AS[i], structure);
real_en = s/i;
}
string = (mis) ?
consens_mis((const char **) AS) : consensus((const char **) AS);
ajFmtPrintF(outf,"%s\n%s", string, structure);
ajFmtPrintF(outf," (%6.2f = %6.2f + %6.2f) \n", min_en, real_en,
min_en-real_en );
if (length<=2500) {
char **A;
A = annote(structure, (const char**) AS);
if (doColor)
(void) PS_rna_plot_a(string, structure, essfile, A[0], A[1]);
else
(void) PS_rna_plot_a(string, structure, essfile, NULL, A[1]);
free(A[0]); free(A[1]);free(A);
} else
ajWarn("INFO: structure too long, not doing xy_plot\n");
if (doAlnPS)
PS_color_aln(structure, alifile, AS, names);
{ /* free mfe arrays but preserve base_pair for PS_dot_plot */
struct bond *bp;
bp = base_pair; base_pair = space(16);
free_alifold_arrays(); /* free's base_pair */
free_alipf_arrays();
base_pair = bp;
}
if (pf) {
double energy, kT;
pair_info *pi;
char * mfe_struc;
mfe_struc = strdup(structure);
kT = (temperature+273.15)*1.98717/1000.; /* in Kcal */
pf_scale = exp(-(sfact*min_en)/kT/length);
if (length>2000)
ajWarn("scaling factor %f\n", pf_scale);
/* init_alipf_fold(length); */
if (confile)
strncpy(structure, ajStrGetPtr(constring), length+1);
energy = (circ) ? alipf_circ_fold(AS, structure, &pi) : alipf_fold(AS, structure, &pi);
if (do_backtrack) {
ajFmtPrintF(outf,"%s", structure);
ajFmtPrintF(outf," [%6.2f]\n", energy);
}
if ((istty)||(!do_backtrack))
ajFmtPrintF(outf," free energy of ensemble = %6.2f kcal/mol\n", energy);
ajFmtPrintF(outf," frequency of mfe structure in ensemble %g\n",
exp((energy-min_en)/kT));
if (do_backtrack) {
FILE *aliout;
cpair *cp;
short *ptable; int k;
ptable = make_pair_table(mfe_struc);
ajFmtPrintF(outf,"\n# Alignment section\n\n");
aliout = ajFileGetFileptr(outf);
fprintf(aliout, "%d sequences; length of alignment %d\n",
n_seq, length);
fprintf(aliout, "alifold output\n");
for (k=0; pi[k].i>0; k++) {
pi[k].comp = (ptable[pi[k].i] == pi[k].j) ? 1:0;
print_pi(pi[k], aliout);
}
fprintf(aliout, "%s\n", structure);
free(ptable);
cp = make_color_pinfo(pi);
(void) PS_color_dot_plot(string, cp, dotfile);
free(cp);
free(mfe_struc);
free(pi);
}
}
if (cstruc!=NULL) free(cstruc);
free(base_pair);
(void) fflush(stdout);
free(string);
free(structure);
for (i=0; AS[i]; i++) {
free(AS[i]); free(names[i]);
}
ajSeqsetDel(&seq);
ajStrDel(&constring);
ajStrDel(&eenergy);
ajStrDel(&edangles);
ajStrDel(&ensbases);
ajFileClose(&confile);
ajFileClose(¶mfile);
ajFileClose(&outf);
ajFileClose(&essfile);
ajFileClose(&alifile);
ajFileClose(&dotfile);
embExit();
return 0;
}
void RNAFuncs::drawRNAAlignment(const string &structure, const string &altStructure, const string &seq1, const string &seq2, const string &strname1, const string &strname2, const string &filename_prefix, const bool atX, const SquigglePlotOptions &options)
{
const double base_fontsize=12;
string base,structname;
float *X, *Y,min_X=0,max_X=0,min_Y=0,max_Y=0;
Uint i;
short *pair_table,*alt_pair_table;
int id_PS,id;
int ps_color_black,ps_color_red,ps_color_blue,ps_color_green;
Uint basenr_x=0, basenr_y=0;
double xpos,ypos;
char buf[10];
bool isDel,isIns;
string filename;
#ifdef HAVE_LIBGD
int id_PNG=0,id_JPG=0;
int png_color_black,png_color_red,png_color_blue,png_color_green;
int jpg_color_black,jpg_color_red,jpg_color_blue,jpg_color_green;
#endif
X = new float[structure.size()];
Y = new float[structure.size()];
assert(seq1.size() == seq2.size());
assert(seq1.size() == structure.size());
assert(seq1.size() == altStructure.size());
pair_table = make_pair_table(structure.c_str());
alt_pair_table = make_pair_table(altStructure.c_str());
i = naview_xy_coordinates(pair_table, X, Y);
if(i!=structure.size())
cerr << "strange things happening in squigglePlot ..." << endl;
// scale image
for(i=0;i<structure.size();i++)
{
X[i]*=static_cast<float>(options.scale);
Y[i]*=static_cast<float>(options.scale);
}
// calculate image dimensions
for(i=0;i<structure.size();i++)
{
min_X=min(min_X,X[i]);
max_X=max(max_X,X[i]);
min_Y=min(min_Y,Y[i]);
max_Y=max(max_Y,Y[i]);
}
// add a border to image size
min_X-=10;
max_X+=10;
min_Y-=10;
max_Y+=10;
//id_PS = g2_open_PS("ali.ps", g2_A4, g2_PS_port);
filename=filename_prefix + ".ps";
id_PS = g2_open_EPSF((char*)filename.c_str());
g2_set_coordinate_system(id_PS,-min_X,-min_Y,1,1);
#ifdef HAVE_LIBGD
if(options.generatePNG)
{
filename=filename_prefix + ".png";
id_PNG=g2_open_gd((char*)filename.c_str(),(int)(max_X-min_X),(int)(max_Y-min_Y),g2_gd_png);
g2_set_coordinate_system(id_PNG,-min_X,-min_Y,1,1);
}
if(options.generateJPG)
{
filename=filename_prefix + ".jpg";
id_JPG=g2_open_gd((char*)filename.c_str(),(int)(max_X-min_X),(int)(max_Y-min_Y),g2_gd_jpeg);
g2_set_coordinate_system(id_PS,-min_X,-min_Y,1,1);
}
#endif
id = g2_open_vd();
g2_attach(id,id_PS);
#ifdef HAVE_LIBGD
if(options.generatePNG)
g2_attach(id,id_PNG);
if(options.generateJPG)
g2_attach(id,id_JPG);
#endif
// cout << "min_X: " << min_X <<",max_X: " << max_X << ",min_Y: " << min_Y << "max_Y: " << max_Y << endl;
// g2_set_coordinate_system(id_PS,595/2.0,842/2.0,0.5,0.5);
g2_set_line_width(id,0.2);
// mark 5' end
g2_string(id,X[0]-20,Y[0],"5'");
// define colors
if(options.greyColors)
{
ps_color_black=g2_ink(id_PS,0,0,0);
ps_color_red=g2_ink(id_PS,0,0,0);
ps_color_blue=g2_ink(id_PS,0.7,0.7,0.7);
ps_color_green=g2_ink(id_PS,0.4,0.4,0.4);
#ifdef HAVE_LIBGD
if(options.generatePNG)
{
png_color_black=g2_ink(id_PNG,0,0,0);
png_color_red=g2_ink(id_PNG,0,0,0);
png_color_blue=g2_ink(id_PNG,0.7,0.7,0.7);
png_color_green=g2_ink(id_PNG,0.4,0.4,0.4);
}
if(options.generateJPG)
{
jpg_color_black=g2_ink(id_JPG,0,0,0);
jpg_color_red=g2_ink(id_JPG,0,0,0);
jpg_color_blue=g2_ink(id_JPG,0.7,0.7,0.7);
jpg_color_green=g2_ink(id_JPG,0.4,0.4,0.4);
}
#endif
}
else
{
ps_color_black=g2_ink(id_PS,0,0,0);
ps_color_red=g2_ink(id_PS,1,0,0);
ps_color_blue=g2_ink(id_PS,0,0,0.5);
ps_color_green=g2_ink(id_PS,0,0.5,0);
#ifdef HAVE_LIBGD
if(options.generatePNG)
{
png_color_black=g2_ink(id_PNG,0,0,0);
png_color_red=g2_ink(id_PNG,1,0,0);
png_color_blue=g2_ink(id_PNG,0,0,0.5);
png_color_green=g2_ink(id_PNG,0,0.5,0);
}
if(options.generateJPG)
{
jpg_color_black=g2_ink(id_JPG,0,0,0);
jpg_color_red=g2_ink(id_JPG,1,0,0);
jpg_color_blue=g2_ink(id_JPG,0,0,0.5);
jpg_color_green=g2_ink(id_JPG,0,0.5,0);
}
#endif
}
// draw sequence
g2_set_font_size(id,base_fontsize*options.scale);
for(i=0;i<structure.size();i++)
{
isDel = false;
isIns = false;
//base
g2_pen(id,ps_color_black); // match
base = "";
if(seq1[i]!='-')
{
base += seq1[i];
basenr_x++;
}
else
{
g2_pen(id,ps_color_green); // insertion
isIns=true;
}
if(seq2[i]!='-')
{
base += seq2[i];
basenr_y++;
}
else
{
g2_pen(id,ps_color_blue); // deletion
isDel=true;
}
if(base.size()==2 && base[0]!=base[1])
g2_pen(id,ps_color_red); // mismatch
else
base = base[0]; // show duplicate base only once
xpos=X[i]-base.length()*(base_fontsize*options.scale)/2;
ypos=Y[i]-4;
g2_string(id,xpos,ypos,(char*)base.c_str());
if(!options.hideBaseNumbers)
{
// draw base number
if(!isIns && basenr_x % options.baseNumInterval == 0)
{
g2_pen(id,ps_color_blue);
sprintf(buf,"%d",basenr_x);
g2_string(id,xpos-20,ypos,buf);
}
if(!isDel && basenr_y % options.baseNumInterval == 0)
{
g2_pen(id,ps_color_green);
sprintf(buf,"%d",basenr_y);
g2_string(id,xpos+20,ypos,buf);
}
}
// connection to next base
if(i<structure.size()-1)
{
// if the connected bases are only in one of the structures
// use the appropriate color
if(seq1[i]=='-' && seq1[i+1]=='-')
g2_pen(id,ps_color_green);
else
if(seq2[i]=='-' && seq2[i+1]=='-')
g2_pen(id,ps_color_blue);
else
g2_pen(id,ps_color_black);
g2_line(id,X[i],Y[i],X[i+1],Y[i+1]);
// draw circles at line endpoints
if(seq1[i]=='-')
g2_pen(id,ps_color_green);
else
if(seq2[i]=='-')
g2_pen(id,ps_color_blue);
else
g2_pen(id,ps_color_black);
g2_filled_circle(id,X[i],Y[i],0.7*options.scale); // circles are drawn twice, but thats ok ...
if(seq1[i+1]=='-')
g2_pen(id,ps_color_green);
else
if(seq2[i+1]=='-')
g2_pen(id,ps_color_blue);
else
g2_pen(id,ps_color_black);
g2_filled_circle(id,X[i+1],Y[i+1],0.7*options.scale);
}
}
// draw pairings
// !!! pair_table indexing begins at 1 !!!
for(i=0;i<structure.size();i++)
{
if((unsigned short)pair_table[i+1]>i+1 && (unsigned short)alt_pair_table[i+1]>i+1)
{
// pairs in both structures
g2_pen(id,ps_color_black);
g2_line(id,X[i],Y[i],X[pair_table[i+1]-1],Y[pair_table[i+1]-1]);
}
else
{
if((unsigned short)pair_table[i+1]>i+1 && (unsigned short)alt_pair_table[i+1]<=i+1)
{
// pairs only in first structure
if(atX)
g2_pen(id,ps_color_blue);
else
g2_pen(id,ps_color_green);
g2_line(id,X[i],Y[i],X[pair_table[i+1]-1],Y[pair_table[i+1]-1]);
}
else
{
if((unsigned short)pair_table[i+1]<=i+1 && (unsigned short)alt_pair_table[i+1]>i+1)
{
// pairs only in second structure
if(atX)
g2_pen(id,ps_color_green);
else
g2_pen(id,ps_color_blue);
double dashes=2.0;
g2_set_dash(id,1,&dashes);
g2_line(id,X[i],Y[i],X[alt_pair_table[i+1]-1],Y[alt_pair_table[i+1]-1]);
dashes=0;
g2_set_dash(id,1,&dashes);
}
}
}
}
// draw structure names
// x-at-y or y-at-x
if(atX)
{
g2_pen(id,ps_color_green);
g2_string(id,min_X,max_Y-10,(char*)strname2.c_str());
g2_pen(id,ps_color_black);
g2_string(id,min_X,max_Y-20,"at");
g2_pen(id,ps_color_blue);
g2_string(id,min_X,max_Y-30,(char*)strname1.c_str());
}
else
{
g2_pen(id,ps_color_blue);
g2_string(id,min_X,max_Y-10,(char*)strname1.c_str());
g2_pen(id,ps_color_black);
g2_string(id,min_X,max_Y-20,"at");
g2_pen(id,ps_color_green);
g2_string(id,min_X,max_Y-30,(char*)strname2.c_str());
}
g2_flush(id);
g2_close(id);
free(pair_table);
free(alt_pair_table);
DELETE(X);
DELETE(Y);
}
int PS_color_aln(const char *structure, const char *filename,
const char *seqs[], const char *names[]) {
/* produce PS sequence alignment color-annotated by consensus structure */
int N,i,j,k,x,y,tmp,columnWidth;
char *tmpBuffer,*ssEscaped,*ruler, *cons;
char c;
float fontWidth, fontHeight, imageHeight, imageWidth,tmpColumns;
int length, maxName, maxNum, currPos;
float lineStep,blockStep,consStep,ssStep,rulerStep,nameStep,numberStep;
float maxConsBar,startY,namesX,seqsX, currY;
float score,barHeight,xx,yy;
int match,block;
FILE *outfile;
short *pair_table;
char * colorMatrix[6][3] = {
{"0.0 1", "0.0 0.6", "0.0 0.2"}, /* red */
{"0.16 1","0.16 0.6", "0.16 0.2"}, /* ochre */
{"0.32 1","0.32 0.6", "0.32 0.2"}, /* turquoise */
{"0.48 1","0.48 0.6", "0.48 0.2"}, /* green */
{"0.65 1","0.65 0.6", "0.65 0.2"}, /* blue */
{"0.81 1","0.81 0.6", "0.81 0.2"} /* violet */
};
const char *alnPlotHeader =
"%%!PS-Adobe-3.0 EPSF-3.0\n"
"%%%%BoundingBox: %i %i %i %i\n"
"%%%%EndComments\n"
"%% draws Vienna RNA like colored boxes\n"
"/box { %% x1 y1 x2 y2 hue saturation\n"
" gsave\n"
" dup 0.3 mul 1 exch sub sethsbcolor\n"
" exch 3 index sub exch 2 index sub rectfill\n"
" grestore\n"
"} def\n"
"%% draws a box in current color\n"
"/box2 { %% x1 y1 x2 y2\n"
" exch 3 index sub exch 2 index sub rectfill\n"
"} def\n"
"/string { %% (Text) x y\n"
" 6 add\n"
" moveto\n"
" show\n"
"} def\n"
"0 %i translate\n"
"1 -1 scale\n"
"/Courier findfont\n"
"[10 0 0 -10 0 0] makefont setfont\n";
outfile = fopen(filename, "w");
if (outfile == NULL) {
fprintf(stderr, "can't open file %s - not doing alignment plot\n",
filename);
return 0;
}
columnWidth=60; /* Display long alignments in blocks of this size */
fontWidth=6; /* Font metrics */
fontHeight=6.5;
lineStep=fontHeight+2; /* distance between lines */
blockStep=3.5*fontHeight; /* distance between blocks */
consStep=fontHeight*0.5; /* distance between alignment and conservation curve */
ssStep=2; /* distance between secondary structure line and sequences */
rulerStep=2; /* distance between sequences and ruler */
nameStep=3*fontWidth; /* distance between names and sequences */
numberStep=fontWidth; /* distance between sequeces and numbers */
maxConsBar=2.5*fontHeight; /* Height of conservation curve */
startY=2; /* "y origin" */
namesX=fontWidth; /* "x origin" */
/* Number of columns of the alignment */
length=strlen(seqs[0]);
/* Allocate memory for various strings, length*2 is (more than)
enough for all of them */
tmpBuffer = (char *) space((unsigned) length*2);
ssEscaped=(char *) space((unsigned) length*2);
ruler=(char *) space((unsigned) length*2);
pair_table=make_pair_table(structure);
/* Get length of longest name and count sequences in alignment*/
for (i=maxName=N=0; names[i] != NULL; i++) {
N++;
tmp=strlen(names[i]);
if (tmp>maxName) maxName=tmp;
}
/* x-coord. where sequences start */
seqsX=namesX+maxName*fontWidth+nameStep;
/* calculate number of digits of the alignment length */
snprintf(tmpBuffer,length, "%i",length);
maxNum=strlen(tmpBuffer);
/* Calculate bounding box */
tmpColumns=columnWidth;
if (length<columnWidth){
tmpColumns=length;
}
imageWidth=ceil(namesX+(maxName+tmpColumns+maxNum)*fontWidth+2*nameStep+fontWidth+numberStep);
imageHeight=startY+ceil((float)length/columnWidth)*((N+2)*lineStep+blockStep+consStep+ssStep+rulerStep);
/* Write postscript header including correct bounding box */
fprintf(outfile,alnPlotHeader,0,0,(int)imageWidth,(int)imageHeight,(int)imageHeight);
/* Create ruler and secondary structure lines */
i=0;
/* Init all with dots */
for (i=0;i<(length);i++){
ruler[i]='.';
}
i=0;
for (i=0;i<length;i++){
/* Write number every 10th position, leave out block breaks */
if ((i+1)%10==0 && (i+1)%columnWidth!=0){
snprintf(tmpBuffer,length,"%i",i+1);
strncpy(ruler+i,tmpBuffer,strlen(tmpBuffer));
}
}
ruler[length]='\0';
/* Draw color annotation first */
/* Repeat for all pairs */
for (i=1; i<=length; i++) {
if ((j=pair_table[i])>i) {
/* Repeat for open and closing position */
for (k=0;k<2;k++){
int pairings, nonpair, s, col;
int ptype[8] = {0,0,0,0,0,0,0,0};
char *color;
col = (k==0)?i-1:j-1;
block=ceil((float)(col+1)/columnWidth);
xx=seqsX+(col-(block-1)*columnWidth)*fontWidth;
/* Repeat for each sequence */
for (s=pairings=nonpair=0; s<N; s++) {
ptype[BP_pair[ENCODE(seqs[s][i-1])][ENCODE(seqs[s][j-1])]]++;
}
for (pairings=0,s=1; s<=7; s++) {
if (ptype[s]) pairings++;
}
nonpair=ptype[0];
if (nonpair <=2) {
color = colorMatrix[pairings-1][nonpair];
for (s=0; s<N; s++) {
yy=startY+(block-1)*(lineStep*(N+2)+blockStep+consStep+rulerStep)+ssStep*(block)+(s+1)*lineStep;
/* Color according due color information in pi-array, only if base pair is possible */
if (BP_pair[ENCODE(seqs[s][i-1])][ENCODE(seqs[s][j-1])]) {
fprintf(outfile, "%.1f %.1f %.1f %.1f %s box\n",
xx,yy-1,xx+fontWidth,yy+fontHeight+1,color);
}
}
}
}
}
}
free(pair_table);
/* Process rest of the output in blocks of columnWidth */
currY=startY;
currPos=0;
cons = consensus(seqs);
while (currPos<length) {
/* Display secondary structure line */
fprintf(outfile,"0 setgray\n");
strncpy(tmpBuffer,structure+currPos,columnWidth);
tmpBuffer[columnWidth]='\0';
x=0;y=0;
while ((c=tmpBuffer[x])){
if (c=='.'){
ssEscaped[y++]='.';
} else {
ssEscaped[y++]='\\';
ssEscaped[y++]=c;
}
x++;
}
ssEscaped[y]='\0';
fprintf(outfile, "(%s) %.1f %.1f string\n", ssEscaped,seqsX,currY);
currY+=ssStep+lineStep;
/* Display names, sequences and numbers */
for (i=0; i<N; i++) {
strncpy(tmpBuffer,seqs[i]+currPos,columnWidth);
tmpBuffer[columnWidth]='\0';
match=0;
for (j=0;j<(currPos+strlen(tmpBuffer));j++){
if (seqs[i][j] != '-') match++;
}
fprintf(outfile, "(%s) %.1f %.1f string\n", names[i],namesX,currY);
fprintf(outfile, "(%s) %.1f %.1f string\n", tmpBuffer,seqsX,currY);
fprintf(outfile, "(%i) %.1f %.1f string\n", match,seqsX+fontWidth*(strlen(tmpBuffer))+numberStep,currY);
currY+=lineStep;
}
currY+=rulerStep;
strncpy(tmpBuffer,ruler+currPos,columnWidth);
tmpBuffer[columnWidth]='\0';
fprintf(outfile, "(%s) %.1f %.1f string\n", tmpBuffer,seqsX,currY);
currY+=lineStep;
currY+=consStep;
/*Display conservation bar*/
fprintf(outfile,"0.6 setgray\n");
for (i=currPos;(i<currPos+columnWidth && i<length);i++){
match=0;
for (j=0;j<N;j++){
if (cons[i] == seqs[j][i]) match++;
if (cons[i]=='U' && seqs[j][i]=='T') match++;
if (cons[i]=='T' && seqs[j][i]=='U') match++;
}
score=(float)(match-1)/(N-1);
if (cons[i] == '-' ||
cons[i] == '_' ||
cons[i] == '.'){
score=0;
}
barHeight=maxConsBar*score;
if (barHeight==0){
barHeight=1;
}
xx=seqsX+(i-(columnWidth*currPos/columnWidth))*fontWidth;
fprintf(outfile,"%.1f %.1f %.1f %.1f box2\n",
xx,
currY+maxConsBar-barHeight,
xx+fontWidth,
currY+maxConsBar);
}
currY+=blockStep;
currPos+=columnWidth;
}
free(cons);
fprintf(outfile,"showpage\n");
fclose(outfile);
free(tmpBuffer);
free(ssEscaped);free(ruler);
return 0;
}
PRIVATE int find_path_once(const char *struc1, const char *struc2, int maxE, int maxl) {
short *pt1, *pt2;
move_t *mlist;
int i, len, d, dist=0, result;
intermediate_t *current, *next;
pt1 = make_pair_table(struc1);
pt2 = make_pair_table(struc2);
len = (int) strlen(struc1);
mlist = (move_t *) space(sizeof(move_t)*len); /* bp_dist < n */
for (i=1; i<=len; i++) {
if (pt1[i] != pt2[i]) {
if (i<pt1[i]) { /* need to delete this pair */
mlist[dist].i = -i;
mlist[dist].j = -pt1[i];
mlist[dist++].when = 0;
}
if (i<pt2[i]) { /* need to insert this pair */
mlist[dist].i = i;
mlist[dist].j = pt2[i];
mlist[dist++].when = 0;
}
}
}
free(pt2);
BP_dist = dist;
current = (intermediate_t *) space(sizeof(intermediate_t)*(maxl+1));
current[0].pt = pt1;
current[0].Sen = current[0].curr_en = energy_of_structure_pt(seq, pt1, S, S1, 0);
current[0].moves = mlist;
next = (intermediate_t *) space(sizeof(intermediate_t)*(dist*maxl+1));
for (d=1; d<=dist; d++) { /* go through the distance classes */
int c, u, num_next=0;
intermediate_t *cc;
for (c=0; current[c].pt != NULL; c++) {
num_next += try_moves(current[c], maxE, next+num_next, d);
}
if (num_next==0) {
for (cc=current; cc->pt != NULL; cc++) free_intermediate(cc);
current[0].Sen=INT_MAX;
break;
}
/* remove duplicates via sort|uniq
if this becomes a bottleneck we can use a hash instead */
qsort(next, num_next, sizeof(intermediate_t),compare_ptable);
for (u=0,c=1; c<num_next; c++) {
if (memcmp(next[u].pt,next[c].pt,sizeof(short)*len)!=0) {
next[++u] = next[c];
} else {
free_intermediate(next+c);
}
}
num_next = u+1;
qsort(next, num_next, sizeof(intermediate_t),compare_energy);
/* free the old stuff */
for (cc=current; cc->pt != NULL; cc++) free_intermediate(cc);
for (u=0; u<maxl && u<num_next; u++) {
current[u] = next[u];
}
for (; u<num_next; u++)
free_intermediate(next+u);
num_next=0;
}
free(next);
path = current[0].moves;
result = current[0].Sen;
free(current[0].pt); free(current);
return(result);
}
void DSU::FindNumbers(int begin, int end, path_t *path, vector<int> &lm_numbers, bool shifts, bool noLP, bool debug, bool no_new)
{
// first resolve small case:
if (end-begin<4) {
bool begins = true;
for (int i=begin+1; i<end; i++) {
// get the minimum
short *tmp_str = make_pair_table(path[i].s);
int tmp_en;
if (pknots) {
Structure str(seq, tmp_str, s0, s1);
tmp_en = move_gradient_pk(seq, &str, s0, s1, shifts, 0);
copy_arr(tmp_str, str.str);
} else {
tmp_en = move_gradient(seq, tmp_str, s0, s1, 0, shifts, noLP);
}
// speedup
if (lm_numbers[begin] != -1 && begins && tmp_en == LM[lm_numbers[begin]].energy && str_eq(LM[lm_numbers[begin]].structure, tmp_str)) {
lm_numbers[i] = lm_numbers[begin];
} else {
begins = false;
if (lm_numbers[end] != -1 && tmp_en == LM[lm_numbers[end]].energy && str_eq(LM[lm_numbers[end]].structure, tmp_str)) {
lm_numbers[i] = lm_numbers[end];
}
}
if (lm_numbers[i]==-1) {
lm_numbers[i] = FindNum(tmp_en, tmp_str);
// update UBlist
if (lm_numbers[i]==-1 && !no_new) {
if (gl_maxen < tmp_en) {
//fprintf(stderr, "exceeds en.: %s %6.2f\n", pt_to_str(tmp_str).c_str(), tmp_en/100.0);
lm_numbers[i] = AddLMtoTBD(tmp_str, tmp_en, EE_DSU, debug);
} else {
if (debug) fprintf(stderr, "cannot find: %s %6.2f\n", pt_to_str(tmp_str).c_str(), tmp_en/100.0);
// add to list of minima and count with them later...
lm_numbers[i] = AddLMtoTBD(tmp_str, tmp_en, NORM_CF, debug);
}
}
} else debug_c++;
free(tmp_str);
}
return ;
}
// da middle one
int pivot = (end+begin)/2;
short *tmp_str = make_pair_table(path[pivot].s);
//fprintf(stderr, "%s\n", pt_to_str(tmp_str).c_str());
int tmp_en = move_gradient(seq, tmp_str, s0, s1, 0, shifts, noLP);
//fprintf(stderr, "%s\n", pt_to_str(tmp_str).c_str());
// speed up:
if (lm_numbers[begin] != -1 && tmp_en == LM[lm_numbers[begin]].energy && str_eq(LM[lm_numbers[begin]].structure, tmp_str)) {
lm_numbers[pivot] = lm_numbers[begin];
} else {
if (lm_numbers[end] != -1 && tmp_en == LM[lm_numbers[end]].energy && str_eq(LM[lm_numbers[end]].structure, tmp_str)) {
lm_numbers[pivot] = lm_numbers[end];
}
}
// normal behaviour
if (lm_numbers[pivot]==-1) {
lm_numbers[pivot] = FindNum(tmp_en, tmp_str);
// update UBlist
if (lm_numbers[pivot]==-1 && !no_new) {
if (gl_maxen < tmp_en) {
//fprintf(stderr, "exceeds en.: %s %6.2f\n", pt_to_str(tmp_str).c_str(), tmp_en/100.0);
lm_numbers[pivot] = AddLMtoTBD(tmp_str, tmp_en, EE_DSU, debug);
} else {
if (debug) fprintf(stderr, "cannot find: %s %6.2f\n", pt_to_str(tmp_str).c_str(), tmp_en/100.0);
// add to list of minima and count with them later...
lm_numbers[pivot] = AddLMtoTBD(tmp_str, tmp_en, NORM_CF, debug);
}
}
} else debug_c++;
free(tmp_str);
// continue recursion:
if ((lm_numbers[pivot]!=lm_numbers[begin] || lm_numbers[pivot]==-1) && pivot-begin>1) FindNumbers(begin, pivot, path, lm_numbers, shifts, noLP, debug, no_new);
if ((lm_numbers[pivot]!=lm_numbers[end] || lm_numbers[pivot]==-1) && end-pivot>1) FindNumbers(pivot, end, path, lm_numbers, shifts, noLP, debug, no_new);
// return maximal energy
return ;
}
