sint palign1(void) /* a profile alignment */
{
sint i,j,temp;
sint entries;
sint *aligned, *group;
float dscore;
lint score;
info("Start of Initial Alignment");
/* calculate sequence weights according to branch lengths of the tree -
weights in global variable seq_weight normalised to sum to INT_SCALE_FACTOR */
temp = INT_SCALE_FACTOR/nseqs;
for (i=0;i<nseqs;i++) seq_weight[i] = temp;
distance_tree = FALSE;
/* do the initial alignment......... */
group = (sint *)ckalloc( (nseqs+1) * sizeof (sint));
for(i=1; i<=profile1_nseqs; ++i)
group[i] = 1;
for(i=profile1_nseqs+1; i<=nseqs; ++i)
group[i] = 2;
entries = nseqs;
aligned = (sint *)ckalloc( (nseqs+1) * sizeof (sint) );
for (i=1;i<=nseqs;i++) aligned[i] = 1;
score = prfalign(group, aligned);
info("Sequences:%d Score:%d",(pint)entries,(pint)score);
group=ckfree((void *)group);
aligned=ckfree((void *)aligned);
for (i=1;i<=nseqs;i++) {
for (j=i+1;j<=nseqs;j++) {
dscore = countid(i,j);
tmat[i][j] = ((double)100.0 - (double)dscore)/(double)100.0;
tmat[j][i] = tmat[i][j];
}
}
return(nseqs);
}
int main(int argc,char **argv)
{
sint i,j,k,n,s;
sint status;
sint result_type;
char c;
char infile[FILENAMELEN+1];
char treefile[FILENAMELEN+1];
float meanid;
FILE *tree;
sint maxres,*gapptr=NULL;
IN_TREEPTR itree;
double dscore;
double meanscore;
double **tmat;
sint window;
sint block_cutoff;
OPT opt;
if(argc!=2) {
fprintf(stdout,"Usage: %s input_aln\n",argv[0]);
exit(1);
}
strcpy(infile,argv[1]);
init_options(&opt);
/* read in the sequences */
seq_input(infile,opt.explicit_type,FALSE,&mult_aln);
if(mult_aln.nseqs<=0) {
error("No sequences in %s\n",infile);
exit(1);
}
window=8;
/* count pairwise residue percent identities */
tmat = (double **) ckalloc( (mult_aln.nseqs+1) * sizeof (double *) );
for(i=0;i<mult_aln.nseqs;i++)
tmat[i] = (double *)ckalloc( (mult_aln.nseqs+1) * sizeof (double) );
meanscore=0;
for (i=0,n=0;i<mult_aln.nseqs;i++) {
for (j=i+1;j<mult_aln.nseqs;j++) {
dscore = countid(mult_aln.seqs[i],mult_aln.seqs[j]);
tmat[j][i] = tmat[i][j] = (100.0 - dscore)/100.0;
n++;
meanscore+=dscore;
}
}
meanscore/=(float)n;
/*if(mult_aln.nseqs<100) block_cutoff=8;
else if(mult_aln.nseqs<250) block_cutoff=6;
else*/ block_cutoff=5;
if(meanscore>50) block_cutoff=50;
/* make a tree from the percent identities (used for sequence weighting) */
strcpy(treefile,infile);
strcat(treefile,".ph");
if((tree = open_explicit_file(treefile))==NULL) exit(1);
guide_tree(tree,mult_aln.seqs,mult_aln.nseqs, tmat, QUICKNJ);
itree=(IN_TREEPTR)ckalloc(sizeof(IN_TREE));
status = read_tree(treefile, mult_aln.seqs, 0, mult_aln.nseqs,itree);
for(i=0;i<mult_aln.nseqs;i++)
ckfree(tmat[i]);
ckfree(tmat);
if (status < 0) exit(1);
seq_weight = calc_seq_weights(0,mult_aln.nseqs,itree,FALSE);
free_tree(itree);
remove(treefile);
/* find the start and end positions of each sequence */
is = (sint *)ckalloc((mult_aln.nseqs+1) * sizeof(sint));
ie = (sint *)ckalloc((mult_aln.nseqs+1) * sizeof(sint));
for(s=0;s<mult_aln.nseqs;s++) {
is[s]=0;
ie[s] = mult_aln.seqs[s].len;
for (i=0; i<mult_aln.seqs[s].len; i++) {
c = mult_aln.seqs[s].data[i];
if (!isalpha(c))
is[s]++;
else
break;
}
for (i=mult_aln.seqs[s].len-1; i>=0; i--) {
c = mult_aln.seqs[s].data[i];
if (!isalpha(c))
ie[s]--;
else
break;
}
}
matrix.format=0;
maxres = get_cl_matrix(FALSE, gon250mt, gapptr, TRUE, 100, &matrix);
all_blocks(infile,window,block_cutoff);
}