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;
}
Alignment * extract_domain ( Constraint_list *CL)
{
/*
function documentation: start
Alignment * extract_domain ( Constraint_list *CL)
given a CL, this function extracts the next best scoring local multiple alignment
It returns a CL where the aligned residues have been indicated in (CL->moca)->forbiden_residues;
the local alignment is extracted with the dp function indicated by
CL->dp_mode: (gotoh_sw_pair_wise)
Evaluation:
CL->get_dp_cost=slow_get_dp_cost;
CL->evaluate_residue_pair=sw_residue_pair_extended_list;
Continuation:
(CL->moca)->evaluate_domain=evaluate_moca_domain;
Cache of CL:
(CL->moca)->cache_cl_with_domain=cache_cl_with_moca_domain;
Domain post processing:
(CL->moca)->make_nol_aln=make_moca_nol_aln;
function documentation: end
*/
int min_start, max_start, start,min_len, max_len, len, score;
int step;
Alignment *C=NULL;
Alignment *RESULT=NULL;
Alignment *EA=NULL;
/*CASE 1: Non Automatic Domain Extraction*/
if ((CL->moca)->moca_interactive)
{
return interactive_domain_extraction (CL);
}
else if ((CL->moca)->moca_len)
{
while ((C=extract_domain_with_coordinates (C,(CL->moca)->moca_start,(CL->moca)->moca_len,CL))->nseq==0)(CL->moca)->moca_scale=(CL->moca)->moca_scale*0.9;
RESULT=copy_aln ( C, RESULT);
unpack_seq_aln (RESULT, CL);
output_format_aln ("mocca_aln",RESULT,EA=fast_coffee_evaluate_output(RESULT, CL),"stdout");
free_aln(EA);
return RESULT;
}
else if ( !(CL->moca)->moca_len)
{
analyse_sequence (CL);
myexit (EXIT_FAILURE);
}
/*CASE 2: Automatic Domain Extraction: Find Coordinates*/
start=500;
step=10;
min_start=0;
max_start=strlen ((CL->S)->seq[0]);
min_len=20;
max_len=strlen ((CL->S)->seq[0]);
C=extract_domain_with_coordinates (C,13,30,CL);
C->output_res_num=1;
print_aln (C);
(CL->moca)->moca_scale=-180;
C=add_seq2aln (CL,C, CL->S);
print_aln (C);
(CL->moca)->moca_scale=-160;
C=add_seq2aln (CL,C, CL->S);
print_aln (C);
myexit (EXIT_FAILURE);
while ( step>0)
{
C=approximate_domain (min_start,max_start,step,min_len,max_len, step,&start, &len, &score, CL);
min_start=start-step;
max_start=start+step;
min_len=len-step;
max_len=len+step;
step=step/2;
}
C=extract_domain_with_coordinates (C,start-10, len+20,CL);
C->output_res_num=1;
print_aln (C);
myexit (EXIT_FAILURE);
return C;
}