char *consensus(const char *AS[]) {
/* simple consensus sequence (most frequent character) */
char *string;
int i,n;
n = strlen(AS[0]);
string = (char *) space((n+1)*sizeof(char));
for (i=0; i<n; i++) {
int s,c,fm, freq[8] = {0,0,0,0,0,0,0,0};
for (s=0; AS[s]!=NULL; s++)
freq[encode_char(AS[s][i])]++;
for (s=c=fm=0; s<8; s++) /* find the most frequent char */
if (freq[s]>fm) {c=s, fm=freq[c];}
if (s>4) s++; /* skip T */
string[i]=Law_and_Order[c];
}
return string;
}
void match_word(const DoubleArrayTrieItem *dat, const ucs4_t * word,
int *match_pos, int *id, int limit)
{
int i, j, p;
for (i = 0,p = 0; word[p] && (limit == 0 || p < limit) && dat[i].base != DATRIE_UNUSED; p ++)
{
int k = encode_char(word[p]);
j = dat[i].base + k;
if (j < 0 || j > DATRIE_SIZE || dat[j].parent != i)
break;
i = j;
}
if (match_pos)
*match_pos = p;
if (id)
*id = i;
}
void vigenere(char* dst, char* src, char* pass, int encode) {
int i;
int j;
int pass_len = strlen(pass);
for (i = 0, j = 0; i < strlen(src); i++, j++) {
if (encode)
dst[i] = encode_char(src[i], pass[j % pass_len]);
else
dst[i] = decode_char(src[i], pass[j % pass_len]);
}
/* Ensure trailing null character since `dst` may not have been initialized
* yet.
*/
dst[i] = '\0';
}
int find_saddle(char *sequence, char *struc1, char *struc2, int max) {
int maxl, maxE, i;
char *tmp;
move_t *bestpath=NULL;
int dir;
maxE = INT_MAX - 1;
seq = sequence;
update_fold_params();
make_pair_matrix();
/* nummerically encode sequence */
S = (short *) space(sizeof(short)*(strlen(seq)+1));
S1 = (short *) space(sizeof(short)*(strlen(seq)+1));
S[0] = S1[0] = (short) strlen(seq);
for (i=0; i< strlen(seq); i++) {
S[i+1] = encode_char(seq[i]);
S1[i+1] = alias[S[i+1]];
}
maxl=1;
do {
int saddleE;
path_fwd = !path_fwd;
if (maxl>max) maxl=max;
saddleE = find_path_once(struc1, struc2, maxE, maxl);
if (saddleE<maxE) {
maxE = saddleE;
if (bestpath) free(bestpath);
bestpath = path;
dir = path_fwd;
} else
free(path);
tmp=struc1; struc1=struc2; struc2=tmp;
maxl *=2;
} while (maxl<2*max);
free(S); free(S1);
path=bestpath;
path_fwd = dir;
return maxE;
}
PRIVATE void make_pairset(void)
{
int i,j;
int sym[MAXALPHA];
make_pair_matrix();
base = strlen(symbolset);
for (i=0; i< base; i++) sym[i] = encode_char(symbolset[i]);
for (i=npairs=0; i< base; i++)
for (j=0; j<base; j++)
if (pair[sym[i]][sym[j]]) {
pairset[npairs++] = symbolset[i];
pairset[npairs++] = symbolset[j];
}
npairs /= 2;
if (npairs==0) nrerror("No pairs in this alphabet!");
}
static void encode_sequence(const char *plain, int len, char *coded) {
assert(len >= 0 && len <= 5);
for (int block = 0; block < 8; block++) {
int octet = get_octet(block);
int junk = get_offset(block);
if (octet >= len) {
pad(&coded[block], 8 - block);
return;
}
char c = shift_right(plain[octet], junk);
if (junk < 0 && octet < len - 1) {
c |= shift_right(plain[octet+1], 8 + junk);
}
coded[block] = encode_char(c);
}
}
PUBLIC float max_matching(char *string, char *structure)
{
int i,j,k,p,t,n1,n2,mm,n;
int max=0, kmax;
int nbp=0, npieces;
int m,s;
short *sq;
/* init */
m = minimum_loopsize;
s = minimum_stacksize;
n = strlen(string);
for ( i=0; i<=n; i++ ) {
for ( j=0; j<=n; j++ ) {
aa[i][j] = 0;
for (t=0; t<=s; t++ ) {
mu[i][j][t] = 0;
}
}
}
n = strlen(string);
sq = (short *) space((unsigned) (n+1)*sizeof(short));
/* encode sequence numerically */
for ( i=1; i<=n; i++ )
sq[i] = encode_char(string[i-1]);
/* interate */
if ( s > 1 ) {
for ( p = m+1; p<=n; p++ ) {
for ( i=1; i<= n-p; i++ ) {
j=i+p;
/* no terminal match */
max = 0; kmax = 0;
for ( k=i; k<j-m; k++ ) {
if( (mu[k][j][s] > 0) ) {
if( (mu[i][k-1][0] + mu[k][j][s]) >= max){
max = mu[i][k-1][0] + mu[k][j][s];
kmax = k;
}
if( (mu[i][k-1][s] + mu[k][j][s]) >= max){
max = mu[i][k-1][s] + mu[k][j][s];
kmax = k;
}
}
if ( mu[i][j-1][0] > max ){
max = mu[i][j-1][0];
kmax = 0;
}
if ( mu[i][j-1][s] > max ){
max = mu[i][j-1][s];
kmax = 0;
}
}
mu[i][j][0] = max;
aa[i][j] = kmax;
/* s or more terminal base pairs */
max = 0;
if ( pair[sq[i]][sq[j]] ) {
if ( mu[i+1][j-1][s-1] > 0 ) {
max = mu[i+1][j-1][s-1];
if ( mu[i+1][j-1][s] > max ) max = mu[i+1][j-1][s];
max = max + NJ_energy(sq[i], sq[j]);
}
}
mu[i][j][s] = max;
/* exactly one terminal base pair */
max = 0;
if ( pair[sq[i]][sq[j]] ) {
max = mu[i+1][j-1][0] + NJ_energy(sq[i], sq[j]);
}
mu[i][j][1] = max;
/* any other number of base pairs */
for( t=2; t <= s-1; t++) {
max = 0;
if ( (pair[sq[i]][sq[j]]) && (mu[i+1][j-1][t-1] > 0) ) {
max = mu[i+1][j-1][t-1] + NJ_energy(sq[i], sq[j]);
}
mu[i][j][t] = max;
}
}
}
max = mu[1][n][0];
if ( mu[1][n][s] >= max ) max = mu[1][n][s];
/* BACKTRACKING */
npieces = 1;
nbp = 0;
pieces[1][1] = 1;
pieces[1][2] = n;
base_pair[0].i = max;
while ( npieces > 0) {
n1 = pieces[npieces][1];
n2 = pieces[npieces][2];
npieces--;
if ( mu[n1][n2][0] >= mu[n1][n2][s] ) { /* n1 ./. n2 */
k = aa[n1][n2];
if ( k > 0 ) { /* pb k...n2 */
if( ((k-1)-n1) > m ) { /*non-vanishing 1st part*/
npieces++ ;
pieces[npieces][1] = n1;
pieces[npieces][2] = k-1;
}
n1 = k;
mm = mu[n1][n2][s];
while ( mu[n1][n2][s] == mm ) {
nbp++;
mm--;
base_pair[nbp].i = n1;
base_pair[nbp].j = n2;
n1++;
n2--;
}
for ( i=s-1; i >=1; i--) {
nbp++;
base_pair[nbp].i = n1;
base_pair[nbp].j = n2;
n1++;
n2--;
}
if( (n2-n1) > m ) {
npieces++;
pieces[npieces][1] = n1;
pieces[npieces][2] = n2;
}
}
else {
if( ((n2-1)-n1) > m ) {
npieces++;
pieces[npieces][1] = n1;
pieces[npieces][2] = n2-1;
}
}
}
else { /* n1...n2 */
mm = mu[n1][n2][s];
while ( mu[n1][n2][s] == mm ) {
nbp++;
mm--;
base_pair[nbp].i = n1;
base_pair[nbp].j = n2;
n1++;
n2--;
}
for ( i=s-1; i >=1; i-- ) {
nbp++;
base_pair[nbp].i = n1;
base_pair[nbp].j = n2;
n1++;
n2--;
}
if( (n2-n1) > m ) {
npieces++;
pieces[npieces][1] = n1;
pieces[npieces][2] = n2;
}
}
} /* REPEAT UNTIL npieces==0 */
} /* END OF THE s>1 BLOCK */
if ( s == 1 ) {
/* iterate */
for ( p=m+1; p<=n; p++ ) {
for( i=1; i<= n-p; i++ ) {
j=i+p;
max = mu[i][j-1][0]; kmax = 0; /* no additional bracket */
for ( k=i; k<j-m; k++ ) {
if ( pair[sq[k]][sq[j]] ) {
if( (mu[i][k-1][0] + mu[k+1][j-1][0]
+ NJ_energy(sq[k], sq[j])) > max ) {
max = mu[i][k-1][0] + mu[k+1][j-1][0]
+ NJ_energy(sq[k], sq[j]);
kmax = k;
}
}
}
mu[i][j][0] = max;
aa[i][j] = kmax;
}
}
max = mu[1][n][0];
/* mm_print_matrix(mu, string); */
/* BACKTRACKING */
base_pair[0].i = max;
npieces = 1;
nbp = 0;
pieces[1][1] = 1;
pieces[1][2] = n;
while ( npieces > 0 ) {
n1 = pieces[npieces][1];
n2 = pieces[npieces][2];
npieces--;
k = aa[n1][n2];
if ( k>0 ) { /* stack von k...n2 */
nbp++;
base_pair[nbp].i = k;
base_pair[nbp].j = n2;
if( (k-n1) > (m+1) ) {
npieces++;
pieces[npieces][1] = n1;
pieces[npieces][2] = k-1;
}
if ( ((n2-1)-(k+1)) > m ) {
npieces++;
pieces[npieces][1] = k+1;
pieces[npieces][2] = n2-1;
}
}
else {
if ( (n2-n1) > (m+1) ) {
npieces++;
pieces[npieces][1] = n1; /* eigentlich unnoetig */
pieces[npieces][2] = n2-1;
}
}
} /* repeat until npieces = 0 */
} /* END OF s==1 */
for ( i=0; i<n; i++ ) structure[i]='.';
structure[n]='\0';
for ( i=1; i<= nbp; i++ ) {
n1 = base_pair[i].i;
n2 = base_pair[i].j;
structure[n1-1] = '(';
structure[n2-1] = ')';
}
return (float)max;
}
/* ------------------------------------------------------------------------- */
PUBLIC void
mm_fill_arrays(char *string, char *structure)
{
int i,j,k,p,t,n;
int max, kmax;
int m,s;
short *sq;
/* init */
m = minimum_loopsize;
s = minimum_stacksize;
n = strlen(string);
for ( i=0; i<=n; i++ ) {
for ( j=0; j<=n; j++ ) {
aa[i][j] = 0;
for (t=0; t<=s; t++ ) {
mu[i][j][t] = 0;
}
}
}
n = strlen(string);
sq = (short *) space((unsigned) (n+1)*sizeof(short));
for ( i=1; i<=n; i++ )
sq[i] = encode_char(string[i-1]);
if ( s > 1 ) { /* case: minimal stack size > 1 */
/* interate */
for ( p = m+1; p<=n; p++ ) {
for ( i=1; i<= n-p; i++ ) {
j = i + p;
/* no terminal match */
max = 0; kmax = 0;
for ( k=i; k<j-m; k++ ) {
if( (mu[k][j][s] > 0) ) {
if( (mu[i][k-1][0] + mu[k][j][s]) >= max){
max = mu[i][k-1][0] + mu[k][j][s];
kmax = k;
}
if( (mu[i][k-1][s] + mu[k][j][s]) >= max){
max = mu[i][k-1][s]+mu[k][j][s];
kmax = k;
}
}
if ( mu[i][j-1][0] > max ){
max = mu[i][j-1][0];
kmax = 0;
}
if ( mu[i][j-1][s] > max ){
max = mu[i][j-1][s];
kmax = 0;
}
}
mu[i][j][0] = max;
aa[i][j] = kmax;
/* s or more terminal base pairs */
max = 0;
if ( pair[sq[i]][sq[j]] ) {
if ( mu[i+1][j-1][s-1] > 0 ) {
max = mu[i+1][j-1][s-1];
if ( mu[i+1][j-1][s] > max ) max = mu[i+1][j-1][s];
max = max + NJ_energy(sq[i], sq[j]);
}
}
mu[i][j][s] = max;
/* exactly one terminal base pair */
max = 0;
if ( pair[sq[i]][sq[j]] ) {
max = mu[i+1][j-1][0] + 1;
}
mu[i][j][1] = max;
/* any other number of base pairs */
for( t=2; t <= s-1; t++) {
max = 0;
if ( (pair[sq[i]][sq[j]]) && (mu[i+1][j-1][t-1] > 0) ) {
max = mu[i+1][j-1][t-1] + NJ_energy(sq[i], sq[j]);
}
mu[i][j][t] = max;
}
}
}
max = mu[1][n][0];
if ( mu[1][n][s] >= max ) max = mu[1][n][s];
}
if ( s == 1 ) { /* case: minimal stack size == 1 */
/* iterate */
for ( p=m+1; p<=n; p++ ) {
for( i=1; i<= n-p; i++ ) {
j = i + p;
max = mu[i][j-1][0]; kmax = 0; /* no additional bracket */
for ( k=i; k<j-m; k++ ) {
if ( pair[sq[k]][sq[j]] ) {
if( (mu[i][k-1][0] + mu[k+1][j-1][0]+1) > max ) {
max = mu[i][k-1][0] + mu[k+1][j-1][0] + NJ_energy(sq[k], sq[j]);
kmax = k;
}
}
}
mu[i][j][0] = max;
aa[i][j] = kmax;
}
}
max = mu[1][n][0];
}
/* mm_print_matrix(mu, string); */
}
PUBLIC void encode_ali_sequence(const char *sequence, short *S, short *s5, short *s3, char *ss, unsigned short *as, int circ){
unsigned int i,l;
unsigned short p;
l = strlen(sequence);
S[0] = (short) l;
s5[0] = s5[1] = 0;
/* make numerical encoding of sequence */
for(i=1; i<=l; i++){
short ctemp;
ctemp=(short) encode_char(toupper(sequence[i-1]));
S[i]= ctemp ;
}
if (oldAliEn){
/* use alignment sequences in all energy evaluations */
ss[0]=sequence[0];
for(i=1; i<l; i++){
s5[i] = S[i-1];
s3[i] = S[i+1];
ss[i] = sequence[i];
as[i] = i;
}
ss[l] = sequence[l];
as[l] = l;
s5[l] = S[l-1];
s3[l] = 0;
S[l+1] = S[1];
s5[1] = 0;
if (circ) {
s5[1] = S[l];
s3[l] = S[1];
ss[l+1] = S[1];
}
}
else{
if(circ){
for(i=l; i>0; i--){
char c5;
c5 = sequence[i-1];
if ((c5=='-')||(c5=='_')||(c5=='~')||(c5=='.')) continue;
s5[1] = S[i];
break;
}
for (i=1; i<=l; i++) {
char c3;
c3 = sequence[i-1];
if ((c3=='-')||(c3=='_')||(c3=='~')||(c3=='.')) continue;
s3[l] = S[i];
break;
}
}
else s5[1]=s3[l]=0;
for(i=1,p=0; i<=l; i++){
char c5;
c5 = sequence[i-1];
if ((c5=='-')||(c5=='_')||(c5=='~')||(c5=='.'))
s5[i+1]=s5[i];
else { /* no gap */
ss[p++]=sequence[i-1]; /*start at 0!!*/
s5[i+1]=S[i];
}
as[i]=p;
}
for (i=l; i>=1; i--) {
char c3;
c3 = sequence[i-1];
if ((c3=='-')||(c3=='_')||(c3=='~')||(c3=='.'))
s3[i-1]=s3[i];
else
s3[i-1]=S[i];
}
}
}
static void ini_ringlist(void) {
int i;
/* needed by function energy_of_struct_pt() from Vienna-RNA-1.4 */
pairList = (short *)calloc(GSV.len + 2, sizeof(short));
assert(pairList != NULL);
typeList = (short *)calloc(GSV.len + 2, sizeof(short));
assert(typeList != NULL);
aliasList = (short *)calloc(GSV.len + 2, sizeof(short));
assert(aliasList != NULL);
pairList[0] = typeList[0] = aliasList[0] = GSV.len;
ptype = (char **)calloc(GSV.len + 2, sizeof(char *));
assert(ptype != NULL);
for (i=0; i<=GSV.len; i++) {
ptype[i] = (char*)calloc(GSV.len + 2, sizeof(char));
assert(ptype[i] != NULL);
}
/* allocate virtual root */
wurzl = (baum *)calloc(1, sizeof(baum));
assert(wurzl != NULL);
/* allocate ringList */
rl = (baum *)calloc(GSV.len+1, sizeof(baum));
assert(rl != NULL);
/* allocate PostOrderList */
/* initialize virtualroot */
wurzl->typ = 'r';
wurzl->nummer = -1;
/* connect virtualroot to ringlist-tree in down direction */
wurzl->down = &rl[GSV.len];
/* initialize post-order list */
make_pair_matrix();
/* initialize rest of ringlist-tree */
for(i = 0; i < GSV.len; i++) {
int c;
GAV.currform[i] = '.';
GAV.prevform[i] = 'x';
pairList[i+1] = 0;
rl[i].typ = 'u';
/* decode base to numeric value */
c = encode_char(GAV.farbe[i]);
rl[i].base = typeList[i+1] = c;
aliasList[i+1] = alias[typeList[i+1]];
/* astablish links for node of the ringlist-tree */
rl[i].nummer = i;
rl[i].next = &rl[i+1];
rl[i].prev = ((i == 0) ? &rl[GSV.len] : &rl[i-1]);
rl[i].up = rl[i].down = NULL;
}
GAV.currform[GSV.len] = GAV.prevform[GSV.len] = '\0';
make_ptypes(aliasList);
rl[i].nummer = i;
rl[i].base = 0;
/* make ringlist circular in next, prev direction */
rl[i].next = &rl[0];
rl[i].prev = &rl[i-1];
/* make virtual basepair for virtualroot */
rl[i].up = wurzl;
rl[i].typ = 'x';
}
