PUBLIC short *
vrna_seq_encode(const char *sequence,
vrna_md_t *md){
unsigned int i, l;
short *S = NULL;
if(sequence && md){
S = vrna_seq_encode_simple(sequence, md);
l = (unsigned int)strlen(sequence);
for(i=1; i<=l; i++)
S[i] = md->alias[S[i]];
S[l+1] = S[1];
S[0] = S[l];
}
return S;
}
PRIVATE void
set_fold_compound(vrna_fold_compound_t *vc,
vrna_md_t *md_p,
unsigned int options,
unsigned int aux){
char *sequence, **sequences;
unsigned int length, s;
int cp; /* cut point for cofold */
char *seq, *seq2;
sequence = NULL;
sequences = NULL;
cp = -1;
/* some default init values */
vc->params = NULL;
vc->exp_params = NULL;
vc->matrices = NULL;
vc->exp_matrices = NULL;
vc->hc = NULL;
vc->auxdata = NULL;
vc->free_auxdata = NULL;
switch(vc->type){
case VRNA_VC_TYPE_SINGLE: sequence = vc->sequence;
seq2 = strdup(sequence);
seq = vrna_cut_point_remove(seq2, &cp); /* splice out the '&' if concatenated sequences and
reset cp... this should also be safe for
single sequences */
vc->cutpoint = cp;
if((cp > 0) && (md_p->min_loop_size == TURN))
md_p->min_loop_size = 0; /* is it safe to set this here? */
free(vc->sequence);
vc->sequence = seq;
vc->length = length = strlen(seq);
vc->sequence_encoding = vrna_seq_encode(seq, md_p);
vc->sequence_encoding2 = vrna_seq_encode_simple(seq, md_p);
if(!(options & VRNA_OPTION_EVAL_ONLY)){
vc->ptype = (aux & WITH_PTYPE) ? vrna_ptypes(vc->sequence_encoding2, md_p) : NULL;
/* backward compatibility ptypes */
vc->ptype_pf_compat = (aux & WITH_PTYPE_COMPAT) ? get_ptypes(vc->sequence_encoding2, md_p, 1) : NULL;
} else {
vc->ptype = NULL;
vc->ptype_pf_compat = NULL;
}
vc->sc = NULL;
free(seq2);
break;
case VRNA_VC_TYPE_ALIGNMENT: sequences = vc->sequences;
vc->length = length = vc->length;
vc->cons_seq = consensus((const char **)sequences);
vc->S_cons = vrna_seq_encode_simple(vc->cons_seq, md_p);
vc->pscore = vrna_alloc(sizeof(int)*((length*(length+1))/2+2));
/* backward compatibility ptypes */
vc->pscore_pf_compat = (aux & WITH_PTYPE_COMPAT) ? vrna_alloc(sizeof(int)*((length*(length+1))/2+2)) : NULL;
oldAliEn = vc->oldAliEn = md_p->oldAliEn;
vc->S = vrna_alloc((vc->n_seq+1) * sizeof(short *));
vc->S5 = vrna_alloc((vc->n_seq+1) * sizeof(short *));
vc->S3 = vrna_alloc((vc->n_seq+1) * sizeof(short *));
vc->a2s = vrna_alloc((vc->n_seq+1) * sizeof(unsigned short *));
vc->Ss = vrna_alloc((vc->n_seq+1) * sizeof(char *));
for (s = 0; s < vc->n_seq; s++) {
vrna_aln_encode(vc->sequences[s],
&(vc->S[s]),
&(vc->S5[s]),
&(vc->S3[s]),
&(vc->Ss[s]),
&(vc->a2s[s]),
md_p);
}
vc->S5[vc->n_seq] = NULL;
vc->S3[vc->n_seq] = NULL;
vc->a2s[vc->n_seq] = NULL;
vc->Ss[vc->n_seq] = NULL;
vc->S[vc->n_seq] = NULL;
vc->scs = NULL;
break;
default: /* do nothing ? */
break;
}
vc->iindx = vrna_idx_row_wise(vc->length);
vc->jindx = vrna_idx_col_wise(vc->length);
/* now come the energy parameters */
add_params(vc, md_p, options);
}
PUBLIC void
vrna_aln_encode(const char *sequence,
short **S_p,
short **s5_p,
short **s3_p,
char **ss_p,
unsigned short **as_p,
vrna_md_t *md){
unsigned int i,l;
unsigned short p;
l = strlen(sequence);
(*s5_p) = (short *) vrna_alloc((l + 2) * sizeof(short));
(*s3_p) = (short *) vrna_alloc((l + 2) * sizeof(short));
(*as_p) = (unsigned short *)vrna_alloc((l + 2) * sizeof(unsigned short));
(*ss_p) = (char *) vrna_alloc((l + 2) * sizeof(char));
/* make numerical encoding of sequence */
(*S_p) = vrna_seq_encode_simple(sequence, md);
(*s5_p)[0] = (*s5_p)[1] = 0;
if(md->oldAliEn){
/* use alignment sequences in all energy evaluations */
(*ss_p)[0]=sequence[0];
for(i=1; i<l; i++){
(*s5_p)[i] = (*S_p)[i-1];
(*s3_p)[i] = (*S_p)[i+1];
(*ss_p)[i] = sequence[i];
(*as_p)[i] = i;
}
(*ss_p)[l] = sequence[l];
(*as_p)[l] = l;
(*s5_p)[l] = (*S_p)[l-1];
(*s3_p)[l] = 0;
(*S_p)[l+1] = (*S_p)[1];
(*s5_p)[1] = 0;
if(md->circ){
(*s5_p)[1] = (*S_p)[l];
(*s3_p)[l] = (*S_p)[1];
(*ss_p)[l+1] = (*S_p)[1];
}
}
else{
if(md->circ){
for(i=l; i>0; i--){
char c5;
c5 = sequence[i-1];
if ((c5=='-')||(c5=='_')||(c5=='~')||(c5=='.')) continue;
(*s5_p)[1] = (*S_p)[i];
break;
}
for (i=1; i<=l; i++) {
char c3;
c3 = sequence[i-1];
if ((c3=='-')||(c3=='_')||(c3=='~')||(c3=='.')) continue;
(*s3_p)[l] = (*S_p)[i];
break;
}
}
else (*s5_p)[1]=(*s3_p)[l]=0;
for(i=1,p=0; i<=l; i++){
char c5;
c5 = sequence[i-1];
if ((c5=='-')||(c5=='_')||(c5=='~')||(c5=='.'))
(*s5_p)[i+1]=(*s5_p)[i];
else { /* no gap */
(*ss_p)[p++]=sequence[i-1]; /*start at 0!!*/
(*s5_p)[i+1]=(*S_p)[i];
}
(*as_p)[i]=p;
}
for (i=l; i>=1; i--) {
char c3;
c3 = sequence[i-1];
if ((c3=='-')||(c3=='_')||(c3=='~')||(c3=='.'))
(*s3_p)[i-1]=(*s3_p)[i];
else
(*s3_p)[i-1]=(*S_p)[i];
}
}
}