main(int argc, char **argv) {
char *aa0;
char libstr[MAX_FN];
char qname[MAX_FN];
int sq0off;
int i, n0;
FILE *fp;
struct pstruct pst, *ppst;
/* stuff from initfa.c/h_init() */
memcpy(qascii,aascii,sizeof(qascii));
/* initialize a pam matrix */
ppst = &pst;
strncpy(ppst->pamfile,"BL50",MAX_FN);
standard_pam(ppst->pamfile,ppst,0,0);
/* this is always protein by default */
ppst->nsq = naa;
ppst->nsqx = naax;
for (i=0; i<=ppst->nsqx; i++) {
ppst->sq[i] = aa[i];
ppst->hsq[i] = haa[i];
ppst->sqx[i]=aax[i]; /* sq = aa */
ppst->hsqx[i]=haax[i]; /* hsq = haa */
}
ppst->sq[ppst->nsqx+1] = ppst->sqx[ppst->nsqx+1] = '\0';
if ((aa0 = calloc(MAXTST,sizeof(char)))==NULL) {
fprintf(stderr,"Cannot allocate aa0\n");
exit(1);
}
initenv(argc, argv, &pst, qname);
alloc_pam(pst.nsq+1,pst.nsq+1, &pst);
initpam2(&pst);
n0 = getseq (qname, qascii, aa0, MAXTST, libstr,&sq0off);
if (!pst.pam_pssm) {
fprintf(stderr," ** ERROR ** No -P PSSM provided\n");
}
else {
ppst->pam2p[0] = alloc_pam2p(n0,pst.nsq);
ppst->pam2p[1] = alloc_pam2p(n0,pst.nsq);
if ((fp = fopen(pst.pgpfile,"rb"))!=NULL) {
read_pssm(aa0, n0, pst.nsq, pst.pamscale,fp,ppst);
}
}
}
int
initpam (char *mfname, struct pstruct *ppst)
{
char line[512], *lp;
int i, j, iaa, pval, p_i, p_j;
int l_nsq;
unsigned char l_sq[MAXSQ+1];
int ess_tmp, max_val, min_val;
int have_es = 0;
FILE *fmat;
pam_opts(mfname, ppst);
if ((fmat = fopen (mfname, "r")) == NULL)
{
printf ("***WARNING*** cannot open scoring matrix file %s\n", mfname);
fprintf (stderr,"***WARNING*** cannot open scoring matrix file %s\n", mfname);
return 0;
}
/* removed because redundant, and causes crash under MacOSX -- because copying on top of itself */
/*
SAFE_STRNCPY (ppst->pamfile, mfname, MAX_FN);
*/
SAFE_STRNCPY(ppst->pam_name, ppst->pamfile, MAX_FN);
if (ppst->pam_ms) {
SAFE_STRNCAT(ppst->pam_name,"-MS",MAX_FN-strlen(ppst->pam_name));
}
/*
the size of the alphabet is determined in advance
*/
ppst->nt_align = (ppst->dnaseq == SEQT_DNA || ppst->dnaseq == SEQT_RNA);
/*
look for alphabet line, skipping the comments, alphabet ends up in line[]
*/
while (fgets (line, sizeof(line), fmat) != NULL && line[0]=='#');
/* transfer the residue line into l_sq[] */
l_nsq = 1;
l_sq[0] = '\0';
for (i=0; i<strlen(line); i++) {
if (isalpha(line[i]) || line[i] == '*') {
l_sq[l_nsq++] = line[i];
}
}
/* if we have a DNA matrix, various defaults must be updated,
particularly pascii, which is used to map the residue ordering
in the matrix file to the residue ordering used by the
program */
if (l_nsq < 20) {
if (ppst->dnaseq <= SEQT_PROT) {
ppst->dnaseq = SEQT_DNA;
}
ppst->nt_align=1;
pascii = nascii; /* use correct DNA mapping, NCBIstdaa by default */
}
/* we no-longer re-initialize sascii[], we either use NCBIstdaa
mapping for protein, or nascii for DNA */
/* 11-July-2014 -- need to check that alphabet is consistent with pascii */
/*
for (i=0; i < l_nsq; i++) {
}
*/
/* check for 2D pam - if not found, allocate it */
if (!ppst->have_pam2) {
alloc_pam (MAXSQ+1, MAXSQ+1, ppst);
ppst->have_pam2 = 1;
}
max_val = -1;
min_val = 1;
ppst->pam2[0][0][0] = -BIGNUM;
/* make certain the [0] boundaries are -BIGNUM */
for (j=1; j < l_nsq; j++) {
p_j = pascii[l_sq[j]];
ppst->pam2[0][0][p_j] = ppst->pam2[0][p_j][0] = -BIGNUM;
}
/* read the scoring matrix values */
for (iaa = 1; iaa < l_nsq; iaa++) { /* read pam value line */
p_i = pascii[l_sq[iaa]];
if (p_i > MAXSQ) {
fprintf(stderr,"*** error [%s:%d] - residue character %c out of range %d\n",
__FILE__, __LINE__, l_sq[iaa], p_i);
p_i = pascii['X'];
}
if (fgets(line,sizeof(line),fmat)==NULL) {
fprintf (stderr," error reading pam line: %s\n",line);
exit (1);
}
/* fprintf(stderr,"%d/%d %s",iaa,nsq,line); */
strtok(line," \t\n"); /* skip the letter (residue) */
for (j = 1; j < l_nsq; j++) {
p_j = pascii[l_sq[j]];
lp=strtok(NULL," \t\n"); /* get the number string */
pval=ppst->pam2[0][p_i][p_j]=atoi(lp); /* convert to integer */
if (pval > max_val) max_val = pval;
if (pval < min_val) min_val = pval;
}
}
ppst->pam_h = max_val;
ppst->pam_l = min_val;
if (ppst->dnaseq==0) {
pam_sq = apam_sq;
pam_sq_n = apam_sq_n;
init_altpam(ppst);
}
else {
pam_sq = npam_sq;
pam_sq_n = npam_sq_n;
}
/* is protein but do not have '*' in alphabet*/
p_i = pascii['*'];
p_j = pascii['X'];
if (!ppst->nt_align && strchr((char *)l_sq,'*')==NULL) {
/* add it */
for (i=0; i< l_nsq; i++) {
ppst->pam2[0][p_i][i] = ppst->pam2[0][p_j][i];
ppst->pam2[0][i][p_i] = ppst->pam2[0][i][p_j];
}
}
/* make sure that X:X is < 0 if -S */
if (ppst->ext_sq_set && ppst->pam2[0][p_j][p_j] >= 0) {
ppst->pam2[0][p_j][p_j] = -1;
}
fclose (fmat);
return 1;
}
