void html_table::emit_colspan (void)
{
cols *b = columns;
cols *c = columns;
int width = 0;
out->put_string("<colgroup>");
while (c != NULL) {
if (b != NULL && b != c && is_gap(b))
/*
* blank column for gap
*/
out->put_string("<col width=\"")
.put_number(is_gap(b))
.put_string("%\" class=\"center\"></col>")
.nl();
width = (get_right(c)*100 + get_effective_linelength()/2)
/ get_effective_linelength()
- (c->left*100 + get_effective_linelength()/2)
/get_effective_linelength();
switch (c->alignment) {
case 'C':
out->put_string("<col width=\"")
.put_number(width)
.put_string("%\" class=\"center\"></col>")
.nl();
break;
case 'R':
out->put_string("<col width=\"")
.put_number(width)
.put_string("%\" class=\"right\"></col>")
.nl();
break;
default:
out->put_string("<col width=\"")
.put_number(width)
.put_string("%\"></col>")
.nl();
}
b = c;
c = c->next;
}
out->put_string("</colgroup>").nl();
}
int precomputed_pair_wise (Alignment *A,int*ns, int **l_s,Constraint_list *CL)
{
int l1, l2, a, b, c;
int nid=0, npos=0, id;
int r1, r2, s1, s2;
l1=strlen(A->seq_al[l_s[0][0]]);
l2=strlen(A->seq_al[l_s[1][0]]);
if (l1!=l2)
{
fprintf ( stderr, "\nERROR: improper use of the function precomputed pairwise:[FATAL:%s]", PROGRAM);
crash ("");
}
else if ( l1==0)
{
A->score_aln=A->score=0;
return 0;
}
for (npos=0, nid=0, a=0; a< ns[0]; a++)
{
s1=l_s[0][a];
for (b=0; b< ns[1]; b++)
{
s2=l_s[1][b];
for ( c=0; c<l1; c++)
{
r1=A->seq_al[s1][c];
r2=A->seq_al[s2][c];
if ( is_gap(r1) || is_gap(r2));
else
{
npos++;
nid+=(r1==r2);
}
}
}
}
id=(npos==0)?0:((nid*100)/npos);
A->score=A->score_aln=id;
return A->score;
}
int html_table::no_gaps (void)
{
int n=0;
cols *c;
for (c = columns; c != NULL; c = c->next)
if (is_gap(c))
n++;
return n;
}
int hasch_seq(char *seq, int **hs, int **lu,int ktup,char *alp)
{
static int a[10];
int i,j,l,limit,code,flag;
char residue;
int alp_lu[10000];
int alp_size;
alp_size=alp[0];
alp++;
for ( i=0; i< alp_size; i++)
{
alp_lu[(int)alp[i]]=i;
}
l=strlen (seq);
limit = (int) pow((double)(alp_size+1),(double)ktup);
hs[0]=(int*)vcalloc ( l+1,sizeof (int));
lu[0]=(int*)vcalloc ( limit+1, sizeof(int));
if ( l==0)myexit(EXIT_FAILURE);
for (i=1;i<=ktup;i++)
a[i] = (int) pow((double)(alp_size+1),(double)(i-1));
for(i=1;i<=(l-ktup+1);++i)
{
code=0;
flag=FALSE;
for(j=1;j<=ktup;++j)
{
if (is_gap(seq[i+j-2])){flag=TRUE;break;}
else residue=alp_lu[(int)seq[i+j-2]];
code+=residue*a[j];
}
if ( flag)continue;
++code;
if (lu[0][code])hs[0][i]=lu[0][code];
lu[0][code]=i;
}
return 0;
}
void aln2hitMat (Alignment *A, char *phitmat)
{
float **ffpHitScoreMatrix;
int i, j, k, l, s;
int nl = A->len_aln;
int inseq = A->nseq;
int itmpScore;
char matrix[100];
char mode[100];
int isim_count, itotal_count, r1, r2;
//Initialization for files
char *pcFileName = A->file[0];
char prefix[200] ={0};
char *hit_matrix_file =(char*) vcalloc(200, sizeof (char));
char *hit_html_file =(char*) vcalloc(200, sizeof (char));
int len = (strrchr(pcFileName,'.')?strrchr(pcFileName,'.')-pcFileName:strlen(pcFileName));
strncpy(prefix, pcFileName, len);
sprintf(hit_matrix_file, "%s%s", prefix, "_aln.hit_matrix");
sprintf(hit_html_file, "%s%s", prefix, ".alnhit_html");
if ( phitmat && strstr ( phitmat, "help"))
aln2hitMat_help();
if(phitmat == NULL) phitmat = (char*) vcalloc(1, sizeof(char)); //such that program could get default value
strget_param (phitmat, "_MODE_", "id", "%s", mode);
strget_param (phitmat, "_MATRIX_", "blosum62mt", "%s", matrix);
fprintf ( stdout, "[START] aln to hit matrix\n");
fprintf ( stdout, " Mode:%s\n", mode);
fprintf ( stdout, " Matrix:%s\n", matrix);
int **mat = read_matrice(matrix);
ffpHitScoreMatrix=(float**)vcalloc (nl, sizeof (float*));
for(i = 0; i < nl; i++)
ffpHitScoreMatrix[i]=(float*)vcalloc (nl-i, sizeof (float));
fprintf (stdout, "Process positions\n", i);
for(i = 0; i < nl; i++)
{
fprintf (stdout, "%d, ", i);
for(j = i; j < nl; j++)
{
if(strm (mode, "id"))
ffpHitScoreMatrix[i][j-i]=generic_get_seq_sim (aln_column2string(A, i), aln_column2string(A, j), (A->cdna_cache)?A->cdna_cache[0]:NULL, matrix);
else if(strm (mode, "pairscore"))
{
isim_count = itotal_count = 0;
for (k=0; k< inseq; k++)
{
r1=tolower(A->seq_al[k][i]);
if (is_gap(r1))continue;
for (l=0; l< inseq; l++)
{
r2=tolower(A->seq_al[l][j]);
if (is_gap (r2))continue;
s=mat[r2-'A'][r1-'A'];
s=(s<=0)?0:1;
isim_count += s;
itotal_count++;
}
}
r1=(isim_count*100)/itotal_count;
ffpHitScoreMatrix[i][j-i] = r1;
}
else
aln2hitMat_help();
}
}
fprintf (stdout, "\n");
output_hit_matrix(hit_matrix_file, ffpHitScoreMatrix, nl);
//Output Hit Score into color html
output_hit_color_html (A, ffpHitScoreMatrix, nl, hit_html_file);
vfree(ffpHitScoreMatrix);
vfree(hit_matrix_file);
vfree(hit_html_file);
fprintf ( stdout, "[END] aln to hit matrix\n");
}
int make_fasta_cdna_pair_wise (Alignment *B,Alignment *A,int*in_ns, int **l_s,Constraint_list *CL, int *diag)
{
int a,c,p,k;
Dp_Result *DPR;
static Dp_Model *M;
int l0, l1;
int len_i, len_j;
int f0=0, f1=0;
int deltaf0, deltaf1, delta;
int nr1, nr2;
int ala, alb, aa0, aa1;
int type;
char **al;
int **tl_s;
int *tns;
/*DEBUG*/
int debug_cdna_fasta=0;
Alignment *DA;
int score;
int state,prev_state;
int t, e;
int a1, a2;
l0=strlen ( B->seq_al[l_s[0][0]]);
l1=strlen ( B->seq_al[l_s[1][0]]);
al=declare_char (2, l0+l1+1);
B=realloc_aln2 (B,B->nseq,l0+l1+1);
free_int (B->cdna_cache, -1);
B->cdna_cache=declare_int(1, l0+l1+1);
if ( !M)M=initialize_dna_dp_model (CL);
M->diag=diag;
tl_s=declare_int (2, 2);tns=vcalloc(2, sizeof(int));tl_s[0][0]=0;tl_s[1][0]=3;tns[0]=tns[1]=1;
DPR=make_fast_dp_pair_wise (A,tns, tl_s,CL,M);
vfree(tns);free_int(tl_s, -1);
/*new_trace_back*/
a=p=0;
aa0=aa1=ala=alb=0;
while ( (k=DPR->traceback[a++])!=M->START);
while ( (k=DPR->traceback[a++])!=M->END)
{
f0=M->model_properties[k][M->F0];
f1=M->model_properties[k][M->F1];
len_i=M->model_properties[k][M->LEN_I];
len_j=M->model_properties[k][M->LEN_J];
type=M->model_properties[k][M->TYPE];
if (type==M->CODING0)
{
deltaf0=(aa0*3+f0)-ala;
deltaf1=(aa1*3+f1)-alb;
delta=MAX(deltaf0, deltaf1);
for (nr1=0, nr2=0,c=0; c<delta; c++, nr1++, nr2++,p++)
{
if (nr1<deltaf0 && ala<l0)al[0][p]=B->seq_al[l_s[0][0]][ala++];
else al[0][p]='-';
if (nr2<deltaf1 && alb<l1)al[1][p]=B->seq_al[l_s[1][0]][alb++];
else al[1][p]='-';
B->cdna_cache[0][p]=M->NON_CODING;
if ( is_gap(al[1][p]) && is_gap(al[0][p]))p--;
else if ( debug_cdna_fasta)fprintf (stderr, "\nUM: %c %c", al[0][p], al[1][p]);
}
for ( c=0; c< 3; c++, p++)
{
if ( c==0)B->cdna_cache[0][p]=M->CODING0;
else if ( c==1)B->cdna_cache[0][p]=M->CODING1;
else if ( c==2)B->cdna_cache[0][p]=M->CODING2;
if (ala<l0)al[0][p]=B->seq_al[l_s[0][0]][ala++];
else al[0][p]='-';
if (alb<l1)al[1][p]=B->seq_al[l_s[1][0]][alb++];
else al[1][p]='-';
if ( is_gap(al[1][p]) && is_gap(al[0][p]))p--;
else if ( debug_cdna_fasta)fprintf (stderr, "\n%d: %c %c",k, al[0][p], al[1][p]);
}
}
aa0+=len_i;
aa1+=len_j;
}
deltaf0=(aa0*3+f0)-ala;
deltaf1=(aa1*3+f1)-alb;
delta=MAX(deltaf0, deltaf1);
for (nr1=0, nr2=0,c=0; c<delta; c++, nr1++, nr2++,p++)
{
if (nr1<deltaf0 && ala<l0)al[0][p]=B->seq_al[l_s[0][0]][ala++];
else al[0][p]='-';
if (nr2<deltaf1 && alb<l1)al[1][p]=B->seq_al[l_s[1][0]][alb++];
else al[1][p]='-';
B->cdna_cache[0][p]=M->NON_CODING;
if ( is_gap(al[1][p]) && is_gap(al[0][p]))p--;
else if ( debug_cdna_fasta)fprintf (stderr, "\nUM: %c %c", al[0][p], al[1][p]);
}
/*End New traceback*/
al[0][p]='\0';
al[1][p]='\0';
sprintf( B->seq_al[l_s[0][0]], "%s", al[0]);
sprintf( B->seq_al[l_s[1][0]], "%s", al[1]);
B->len_aln=strlen (al[0]);
B->nseq=2;
if ( debug_cdna_fasta)
{
fprintf ( stderr, "\nA-A=%d, %d", CL->M['a'-'A']['a'-'A'], CL->M['a'-'A']['a'-'A'] *SCORE_K);
for ( a=1; a<diag[0]; a++)
{
fprintf ( stderr, "\nchosen diag: %d", diag[a]);
}
fprintf ( stderr, "\n GOP=%d GEP=%d TG_MODE=%d", M->gop, M->gep, M->TG_MODE);
fprintf ( stderr, "\nF_GOP=%d F_GEP=%d F_TG_MODE=%d", M->gop, M->gep, M->F_TG_MODE);
DA=copy_aln (B, NULL);
DA=realloc_aln2 (DA,6,(DA->len_aln+1));
for ( a=0; a<B->len_aln; a++)
{
fprintf ( stderr, "\n%d", DA->cdna_cache[0][a]);
if (DA->cdna_cache[0][a]>=M->CODING0)DA->seq_al[DA->nseq][a]=DA->cdna_cache[0][a]-M->nstate+'0';
else DA->seq_al[DA->nseq][a]=DA->cdna_cache[0][a]-M->nstate+'0';
if (DA->cdna_cache[0][a]==M->CODING0)
{
DA->seq_al[DA->nseq+1][a]=translate_dna_codon (DA->seq_al[0]+a,'*');
DA->seq_al[DA->nseq+2][a]=translate_dna_codon (DA->seq_al[1]+a,'*');
}
else
{
DA->seq_al[DA->nseq+1][a]='-';
DA->seq_al[DA->nseq+2][a]='-';
}
}
DA->nseq+=3;
print_aln (DA);
free_aln(DA);
score=0;
for (prev_state=M->START,a=0; a< DA->len_aln;)
{
state=DA->cdna_cache[0][a];
t=M->model[prev_state][state];
if ( DA->cdna_cache[0][a]==M->CODING0)
{
a1=translate_dna_codon (A->seq_al[0]+a,'x');
a2=translate_dna_codon (A->seq_al[1]+a,'x');
if ( a1!='x' && a2!='x')
{
e=CL->M[a1-'A'][a2-'A']*SCORE_K;
}
}
else if ( DA->cdna_cache[0][a]>M->CODING0);
else
{
e=M->model_properties[B->cdna_cache[0][a]][M->EMISSION];
}
if ( e==UNDEFINED || t==UNDEFINED) fprintf ( stderr, "\nPROBLEM %d\n", a);
fprintf ( stderr, "\n[%c..%c: %d(e)+%d(t)=%d]", A->seq_al[0][a], A->seq_al[1][a], e,t,e+t);
score+=e+t;
prev_state=state;
if (B->cdna_cache[0][a]==M->NON_CODING)a++;
else a+=3;
}
}
for ( a=0; a<B->len_aln; a++)
{
if ( B->cdna_cache[0][a]<M->CODING0)B->cdna_cache[0][a]=0;
else B->cdna_cache[0][a]=1;
}
free_char ( al, -1);
return DPR->score;
}
void html_table::emit_col (int n)
{
cols *c = columns;
cols *b = columns;
int width = 0;
// must be a different row
if (last_col != NULL && n <= last_col->no)
emit_new_row();
while (c != NULL && c->no < n)
c = c->next;
// can we find column, n?
if (c != NULL && c->no == n) {
// shutdown previous column
if (last_col != NULL)
out->put_string("</td>").nl();
// find previous column
if (last_col == NULL)
b = columns;
else
b = last_col;
// have we a gap?
if (last_col != NULL) {
emit_td(is_gap(b), "></td>");
b = b->next;
}
// move across to column n
while (b != c) {
// we compute the difference after converting positions
// to avoid rounding errors
width = (get_right(b)*100 + get_effective_linelength()/2)
/ get_effective_linelength()
- (b->left*100 + get_effective_linelength()/2)
/get_effective_linelength();
emit_td(width, "></td>");
// have we a gap?
emit_td(is_gap(b), "></td>");
b = b->next;
}
width = (get_right(b)*100 + get_effective_linelength()/2)
/ get_effective_linelength()
- (b->left*100 + get_effective_linelength()/2)
/get_effective_linelength();
switch (b->alignment) {
case 'C':
emit_td(width, " align=center>");
break;
case 'R':
emit_td(width, " align=right>");
break;
default:
emit_td(width);
}
// remember column, b
last_col = b;
}
}
