void setup_environment(Char *argv[], boolean *canbeplotted)
{
boolean firsttree;
char* treestr;
/* Set up all kinds of fun stuff */
#ifdef MAC
OSErr retcode;
FInfo fndrinfo;
macsetup("Drawgram","Preview");
#endif
#ifdef TURBOC
if ((registerbgidriver(EGAVGA_driver) <0) ||
(registerbgidriver(Herc_driver) <0) ||
(registerbgidriver(CGA_driver) <0)){
printf("Graphics error: %s ",grapherrormsg(graphresult()));
exit(-1);}
#endif
printf("DRAWGRAM from PHYLIP version %s\n",VERSION);
printf("Reading tree ... \n");
firsttree = true;
treestr = ajStrGetuniquePtr(&phylotrees[0]->Tree);
allocate_nodep(&nodep, treestr, &spp);
treeread (&treestr, &root, treenode, &goteof, &firsttree,
nodep, &nextnode, &haslengths,
&grbg, initdrawgramnode,true,-1);
root->oldlen = 0.0;
printf("Tree has been read.\n");
printf("Loading the font .... \n");
loadfont(font,argv[0]);
printf("Font loaded.\n");
previewing = false;
ansi = ANSICRT;
ibmpc = IBMCRT;
firstscreens = true;
initialparms();
(*canbeplotted) = false;
} /* setup_environment */
void maketree()
{
/* constructs a binary tree from the pointers in treenode.
adds each node at location which yields highest "likelihood"
then rearranges the tree for greatest "likelihood" */
long i, j, nextnode;
double gotlike;
node *item, *nufork, *dummy, *p;
char *treestr;
fullset = (1L << (bits + 1)) - (1L << 1);
if (!usertree) {
for (i = 1; i <= (spp); i++)
enterorder[i - 1] = i;
if (jumble)
randumize(seed, enterorder);
root = treenode[enterorder[0] - 1];
add(treenode[enterorder[0] - 1], treenode[enterorder[1] - 1],
treenode[spp], &root, treenode);
if (progress) {
printf("Adding species:\n");
writename(0, 2, enterorder);
#ifdef WIN32
phyFillScreenColor();
#endif
}
lastrearr = false;
for (i = 3; i <= (spp); i++) {
bestyet = -350.0 * spp * chars;
item = treenode[enterorder[i - 1] - 1];
nufork = treenode[spp + i - 2];
addpreorder(root, item, nufork);
add(there, item, nufork, &root, treenode);
like = bestyet;
rearrange(&root);
if (progress) {
writename(i - 1, 1, enterorder);
#ifdef WIN32
phyFillScreenColor();
#endif
}
lastrearr = (i == spp);
if (lastrearr) {
if (progress) {
printf("\nDoing global rearrangements\n");
printf(" !");
for (j = 1; j <= (nonodes); j++)
if ( j % (( nonodes / 72 ) + 1 ) == 0 )
putchar('-');
printf("!\n");
#ifdef WIN32
phyFillScreenColor();
#endif
}
bestlike = bestyet;
if (jumb == 1) {
bstlike2 = bestlike;
nextree = 1;
}
do {
if (progress)
printf(" ");
gotlike = bestlike;
for (j = 0; j < (nonodes); j++) {
bestyet = - 350.0 * spp * chars;
item = treenode[j];
if (item != root) {
nufork = treenode[j]->back;
re_move(&item, &nufork, &root, treenode);
there = root;
addpreorder(root, item, nufork);
add(there, item, nufork, &root, treenode);
}
if (progress) {
if ( j % (( nonodes / 72 ) + 1 ) == 0 )
putchar('.');
fflush(stdout);
}
}
if (progress) {
putchar('\n');
#ifdef WIN32
phyFillScreenColor();
#endif
}
} while (bestlike > gotlike);
}
}
if (progress)
putchar('\n');
for (i = spp - 1; i >= 1; i--)
re_move(&treenode[i], &dummy, &root, treenode);
if (jumb == njumble) {
if (treeprint) {
putc('\n', outfile);
if (nextree == 2)
fprintf(outfile, "One most parsimonious tree found:\n");
else
fprintf(outfile, "%6ld trees in all found\n", nextree - 1);
}
if (nextree > maxtrees + 1) {
if (treeprint)
fprintf(outfile, "here are the first%4ld of them\n", (long)maxtrees);
nextree = maxtrees + 1;
}
if (treeprint)
putc('\n', outfile);
for (i = 0; i <= (nextree - 2); i++) {
root = treenode[0];
add(treenode[0], treenode[1], treenode[spp], &root, treenode);
for (j = 3; j <= spp; j++) {
add(treenode[bestrees[i].btree[j - 1] - 1], treenode[j - 1],
treenode[spp + j - 2], &root, treenode);}
evaluate(root);
printree(1.0, treeprint, root);
describe();
for (j = 1; j < (spp); j++)
re_move(&treenode[j], &dummy, &root, treenode);
}
}
} else {
if (numtrees > 2) {
emboss_initseed(inseed, &inseed0, seed);
printf("\n");
}
if (treeprint) {
fprintf(outfile, "User-defined tree");
if (numtrees > 1)
putc('s', outfile);
fprintf(outfile, ":\n");
}
names = (boolean *)Malloc(spp*sizeof(boolean));
which = 1;
firsttree = true; /**/
nodep = NULL; /**/
nextnode = 0; /**/
haslengths = 0; /**/
phirst = 0; /**/
zeros = (long *)Malloc(chars*sizeof(long)); /**/
for (i = 0; i < chars; i++) /**/
zeros[i] = 0; /**/
while (which <= numtrees) {
treestr = ajStrGetuniquePtr(&phylotrees[which-1]->Tree);
treeread(&treestr, &root, treenode, &goteof, &firsttree,
nodep, &nextnode, &haslengths,
&grbg, initdollopnode,false,nonodes);
for (i = spp; i < (nonodes); i++) {
p = treenode[i];
for (j = 1; j <= 3; j++) {
p->stateone = (bitptr)Malloc(words*sizeof(long));
p->statezero = (bitptr)Malloc(words*sizeof(long));
p = p->next;
}
} /* debug: see comment at initdollopnode() */
if (treeprint)
fprintf(outfile, "\n\n");
evaluate(root);
printree(1.0, treeprint, root);
describe();
which++;
}
FClose(intree);
fprintf(outfile, "\n\n");
if (numtrees > 1 && chars > 1)
standev(numtrees, minwhich, minsteps, nsteps, fsteps, seed);
free(names);
}
if (jumb == njumble) {
if (progress) {
printf("Output written to file \"%s\"\n\n", outfilename);
if (trout)
printf("Trees also written onto file \"%s\"\n\n", outtreename);
}
if (ancseq)
freegarbage(&garbage);
}
} /* maketree */
AjBool ajReadlinePos(AjPFile file, AjPStr* Pdest, ajlong* Ppos)
{
const char *cp;
char *buff;
ajint isize;
ajint ilen;
ajint jlen;
ajint ipos;
ajuint buffsize;
size_t iread;
const char* pnewline = NULL;
MAJSTRDEL(Pdest);
if(file->Buffsize)
buffsize = file->Buffsize;
else
buffsize = ajFileValueBuffsize();
if(!file->Buff)
ajStrAssignResC(&file->Buff, buffsize, "");
else if(buffsize > MAJSTRGETRES(file->Buff))
ajStrSetRes(&file->Buff, buffsize);
if(MAJSTRGETUSE(file->Buff) == 1)
buff = MAJSTRGETPTR(file->Buff);
else
buff = ajStrGetuniquePtr(&file->Buff);
isize = MAJSTRGETRES(file->Buff);
ilen = 0;
ipos = 0;
if(!file->fp)
ajWarn("ajFileGets file not found");
*Ppos = file->Filepos;
while(buff)
{
if(file->End)
{
ajStrAssignClear(Pdest);
ajDebug("at EOF: File already read to end %F\n", file);
return ajFalse;
}
#ifndef __ppc__
if(file->Readblock)
{
if(file->Blockpos >= file->Blocklen)
{
iread = fread(file->Readblock,
1, file->Blocksize,
file->fp);
if(!iread && ferror(file->fp))
ajFatal("fread failed with error:%d '%s'",
ferror(file->fp), strerror(ferror(file->fp)));
file->Blockpos = 0;
file->Blocklen = iread;
file->Readblock[iread] = '\0';
/*ajDebug("++ fread %u Ppos:%Ld\n", iread, *Ppos);*/
}
if(file->Blockpos < file->Blocklen)
{
/* we know we have something in Readblock to process */
pnewline = strchr(&file->Readblock[file->Blockpos], '\n');
if(pnewline)
jlen = pnewline - &file->Readblock[file->Blockpos] + 1;
else
jlen = file->Blocklen - file->Blockpos;
/*ajDebug("ipos:%d jlen:%d pnewline:%p "
"Readblock:%p blockpos:%d blocklen:%d\n",
ipos, jlen, pnewline, file->Readblock,
file->Blockpos, file->Blocklen);*/
memmove(&buff[ipos], &file->Readblock[file->Blockpos], jlen);
buff[ipos+jlen]='\0';
cp = &buff[ipos];
file->Blockpos += jlen;
}
else
{
jlen = 0;
cp = NULL;
}
}
else
{
cp = fgets(&buff[ipos], isize, file->fp);
jlen = strlen(&buff[ipos]);
}
#else
cp = ajSysFuncFgets(&buff[ipos], isize, file->fp);
jlen = strlen(&buff[ipos]);
#endif
if(!cp && !ipos)
{
if(feof(file->fp))
{
file->End = ajTrue;
ajStrAssignClear(Pdest);
ajDebug("EOF ajFileGetsL file %F\n", file);
return ajFalse;
}
else
ajFatal("Error reading from file '%S'\n", ajFileGetNameS(file));
}
ilen += jlen;
file->Filepos += jlen;
/*
** We need to read again if:
** We have read the entire buffer
** and we don't have a newline at the end
** (must be careful about that - we may just have read enough)
*/
if(((file->Readblock && !pnewline) ||(jlen == (isize-1))) &&
(buff[ilen-1] != '\n'))
{
MAJSTRSETVALIDLEN(&file->Buff, ilen); /* fix before resizing! */
ajStrSetResRound(&file->Buff, ilen+buffsize+1);
/*ajDebug("more to do: jlen: %d ipos: %d isize: %d ilen: %d "
"Size: %d\n",
jlen, ipos, isize, ilen, ajStrGetRes(file->Buff));*/
ipos += jlen;
buff = ajStrGetuniquePtr(&file->Buff);
isize = ajStrGetRes(file->Buff) - ipos;
/*ajDebug("expand to: ipos: %d isize: %d Size: %d\n",
ipos, isize, ajStrGetRes(file>Buff));*/
}
else
buff = NULL;
}
MAJSTRSETVALIDLEN(&file->Buff, ilen);
if (ajStrGetCharLast(file->Buff) != '\n')
{
/*ajDebug("Appending missing newline to '%S'\n", file->Buff);*/
ajStrAppendK(&file->Buff, '\n');
}
ajStrAssignRef(Pdest, file->Buff);
/*
if(file->Readblock)
ajDebug("ajFileGetsL done blocklen:%d blockpos:%d readlen:%u\n",
file->Blocklen, file->Blockpos, ajStrGetLen(file->Buff));
*/
return ajTrue;
}
ajint ajUserGet(AjPStr* pthis, const char* fmt, ...)
{
AjPStr thys;
const char *cp;
char *buff;
va_list args;
ajint ipos;
ajint isize;
ajint ilen;
ajint jlen;
ajint fileBuffSize = ajFileValueBuffsize();
va_start(args, fmt);
ajFmtVError(fmt, args);
va_end(args);
if(ajFileValueRedirectStdin())
{
ajUser("(Standard input in use: using default)");
ajStrAssignC(pthis, "");
return ajStrGetLen(*pthis);
}
ajStrSetRes(pthis, fileBuffSize);
buff = ajStrGetuniquePtr(pthis);
thys = *pthis;
isize = ajStrGetRes(thys);
ilen = 0;
ipos = 0;
/*ajDebug("ajUserGet buffer len: %d res: %d ptr: %x\n",
ajStrGetLen(thys), ajStrGetRes(thys), thys->Ptr);*/
while(buff)
{
#ifndef __ppc__
cp = fgets(&buff[ipos], isize, stdin);
#else
cp = ajSysFuncFgets(&buff[ipos], isize, stdin);
#endif
if(!cp && !ipos)
{
if(feof(stdin))
{
ajErr("Unable to get reply from user - end of standard input");
ajExitBad();
}
else
ajFatal("Error reading from user: '******'\n",
strerror(errno));
}
jlen = strlen(&buff[ipos]);
ilen += jlen;
/*
** We need to read again if:
** We have read the entire buffer
** and we don't have a newline at the end
** (must be careful about that - we may just have read enough)
*/
ajStrSetValidLen(pthis, ilen);
thys = *pthis;
if((jlen == (isize-1)) &&
(ajStrGetCharLast(thys) != '\n'))
{
ajStrSetRes(pthis, ajStrGetRes(thys)+fileBuffSize);
thys = *pthis;
/*ajDebug("more to do: jlen: %d ipos: %d isize: %d ilen: %d "
"Size: %d\n",
jlen, ipos, isize, ilen, ajStrGetRes(thys));*/
ipos += jlen;
buff = ajStrGetuniquePtr(pthis);
isize = ajStrGetRes(thys) - ipos;
/* ajDebug("expand to: ipos: %d isize: %d Size: %d\n",
ipos, isize, ajStrGetRes(thys)); */
}
else
buff = NULL;
}
ajStrSetValidLen(pthis, ilen);
if(ajStrGetCharLast(*pthis) == '\n')
ajStrCutEnd(pthis, 1);
/* PC files have \r\n Macintosh files have just \r : this fixes both */
if(ajStrGetCharLast(*pthis) == '\r')
{
/*ajDebug("Remove carriage-return characters from PC-style files\n");*/
ajStrCutEnd(pthis, 1);
}
ajStrTrimWhite(pthis);
return ajStrGetLen(*pthis);
}
static void merger_Merge(AjPAlign align, AjPStr *ms,
const char *a, const char *b,
const AjPStr m, const AjPStr n, ajint start1,
ajint start2,
const char *namea, const char *nameb)
{
ajint apos;
ajint bpos;
ajint i;
AjPStr mm = NULL;
AjPStr nn = NULL;
char *p;
char *q;
ajint olen; /* length of walk alignment */
size_t tt;
/* lengths of the sequences after the aligned region */
ajint alen;
ajint blen;
AjPStr tmpstr = NULL;
mm = ajStrNewS(m);
nn = ajStrNewS(n);
p = ajStrGetuniquePtr(&mm);
q = ajStrGetuniquePtr(&nn);
olen = ajStrGetLen(mm);
/* output the left hand side */
if(start1 > start2)
{
for(i=0; i<start1; i++)
ajStrAppendK(ms, a[i]);
if(start2)
{
ajFmtPrintAppS(&tmpstr, "WARNING: *************************"
"********\n");
ajFmtPrintAppS(&tmpstr, "The region of alignment only starts at "
"position %d of sequence %s\n", start2+1, nameb);
ajFmtPrintAppS(&tmpstr, "Only the sequence of %s is being used "
"before this point\n\n", namea);
ajAlignSetTailApp(align, tmpstr);
ajStrDel(&tmpstr);
}
}
else if(start2 > start1)
{
for(i=0; i<start2; i++)
ajStrAppendK(ms, b[i]);
if(start1)
{
ajFmtPrintAppS(&tmpstr, "WARNING: **************************"
"*******\n");
ajFmtPrintAppS(&tmpstr, "The region of alignment only starts at "
"position %d of sequence %s\n", start1+1, namea);
ajFmtPrintAppS(&tmpstr, "Only the sequence of %s is being used "
"before this point\n\n", nameb);
ajAlignSetTailApp(align, tmpstr);
ajStrDel(&tmpstr);
}
}
else if(start1 && start2)
{
/* both the same length and > 1 before the aligned region */
ajFmtPrintAppS(&tmpstr,
"WARNING: *********************************\n");
ajFmtPrintAppS(&tmpstr, "There is an equal amount of unaligned "
"sequence (%d) at the start of the sequences.\n",
start1);
ajFmtPrintAppS(&tmpstr, "Sequence %s is being arbitrarily chosen "
"for the merged sequence\n\n", namea);
ajAlignSetTailApp(align, tmpstr);
ajStrDel(&tmpstr);
for(i=0; i<start1; i++)
ajStrAppendK(ms, a[i]);
}
/* header */
ajFmtPrintS(&tmpstr, "Conflicts: %15.15s %15.15s\n", namea, nameb);
ajFmtPrintAppS(&tmpstr,
" position base position base Using\n");
ajAlignSetTailApp(align, tmpstr);
/* make the merged sequence
**
** point to the start of the alignment in the complete unaligned
** sequences
*/
apos = start1;
bpos = start2;
for(i=0; i<olen; i++)
{
if(p[i]=='.' || p[i]==' ' || p[i]=='-' ||
q[i]=='.' || q[i]==' ' || q[i]=='-')
{ /* gap! */
if(merger_bestquality(a, b, apos, bpos))
{
p[i] = toupper((ajint)p[i]);
if(p[i] != '.' && p[i] != ' ' && p[i] != '-')
ajStrAppendK(ms, p[i]);
ajFmtPrintS(&tmpstr,
" %8d '%c' %8d '%c' '%c'\n",
apos+1, p[i], bpos+1, q[i], p[i]);
ajAlignSetTailApp(align, tmpstr);
}
else
{
q[i] = toupper((ajint)q[i]);
if(q[i] != '.' && q[i] != ' ' && q[i] != '-')
ajStrAppendK(ms, q[i]);
ajFmtPrintS(&tmpstr,
" %8d '%c' %8d '%c' '%c'\n",
apos+1, p[i], bpos+1, q[i], q[i]);
ajAlignSetTailApp(align, tmpstr);
}
}
else if(p[i]=='n' || p[i]=='N')
{
q[i] = toupper((ajint)q[i]);
ajStrAppendK(ms, q[i]);
}
else if(q[i]=='n' || q[i]=='N')
{
p[i] = toupper((ajint)p[i]);
ajStrAppendK(ms, p[i]);
}
else if(p[i] != q[i])
{
/*
** get the sequence with the best quality and use the base
** from that one
*/
if(merger_bestquality(a, b, apos, bpos))
{
p[i] = toupper((ajint)p[i]);
ajStrAppendK(ms, p[i]);
ajFmtPrintS(&tmpstr,
" %8d '%c' %8d '%c' '%c'\n",
apos+1, p[i], bpos+1, q[i], p[i]);
ajAlignSetTailApp(align, tmpstr);
}
else
{
q[i] = toupper((ajint)q[i]);
ajStrAppendK(ms, q[i]);
ajFmtPrintS(&tmpstr,
" %8d '%c' %8d '%c' '%c'\n",
apos+1, p[i], bpos+1, q[i], q[i]);
ajAlignSetTailApp(align, tmpstr);
}
}
else
ajStrAppendK(ms, p[i]);
/* update the positions in the unaligned complete sequences */
if(p[i] != '.' && p[i] != ' ' && p[i] != '-') apos++;
if(q[i] != '.' && q[i] != ' ' && q[i] != '-') bpos++;
}
/* output the right hand side */
tt = strlen(&a[apos]);
alen = (ajint) tt;
tt = strlen(&b[bpos]);
blen = (ajint) tt;
if(alen > blen)
{
ajStrAppendC(ms, &a[apos]);
if(blen)
{
ajFmtPrintAppS(&tmpstr, "WARNING: ***************************"
"******\n");
ajFmtPrintAppS(&tmpstr, "The region of alignment ends at "
"position %d of sequence %s\n",
bpos+1, nameb);
ajFmtPrintAppS(&tmpstr, "Only the sequence of %s is being used "
"after this point\n\n", namea);
ajAlignSetTailApp(align, tmpstr);
ajStrDel(&tmpstr);
}
}
if(blen > alen)
{
ajStrAppendC(ms, &b[bpos]);
if(alen)
{
ajFmtPrintAppS(&tmpstr, "WARNING: ***************************"
"******\n");
ajFmtPrintAppS(&tmpstr, "The region of alignment ends at "
"position %d of sequence %s\n",
apos+1, namea);
ajFmtPrintAppS(&tmpstr, "Only the sequence of %s is being used "
"after this point\n\n", nameb);
ajAlignSetTailApp(align, tmpstr);
ajStrDel(&tmpstr);
}
}
else if(alen && blen)
{ /* both the same length and > 1 */
ajFmtPrintAppS(&tmpstr, "WARNING: ************************"
"*********\n");
ajFmtPrintAppS(&tmpstr, "There is an equal amount of unaligned "
"sequence (%d) at the end of the sequences.\n",
alen);
ajFmtPrintAppS(&tmpstr, "Sequence %s is being arbitrarily chosen "
"for the merged sequence\n\n", namea);
ajStrAppendC(ms, &a[apos]);
ajAlignSetTailApp(align, tmpstr);
ajStrDel(&tmpstr);
}
ajStrDel(&mm);
ajStrDel(&nn);
ajStrDel(&tmpstr);
return;
}
static size_t dbiblast_memfreadS(AjPStr* dest, size_t size, size_t num_items,
PMemFile mf)
{
return dbiblast_memfread(ajStrGetuniquePtr(dest), size, num_items, mf);
}
void maketree()
{ /* construct the tree */
long i;
char* treestr;
if (usertree) {
if(numtrees > MAXSHIMOTREES)
shimotrees = MAXSHIMOTREES;
else
shimotrees = numtrees;
if (numtrees > 2)
emboss_initseed(inseed, &inseed0, seed);
if (treeprint) {
fprintf(outfile, "User-defined tree");
if (numtrees > 1)
putc('s', outfile);
putc('\n', outfile);
}
setuptree(&curtree, nonodes2);
for (which = 1; which <= spp; which++)
inittip(which, &curtree);
which = 1;
while (which <= numtrees) {
treestr = ajStrGetuniquePtr(&phylotrees[which-1]->Tree);
treeread2 (&treestr, &curtree.start, curtree.nodep,
lengths, &trweight, &goteof, &haslengths, &spp,false,nonodes2);
curtree.start = curtree.nodep[outgrno - 1]->back;
treevaluate();
printree();
summarize();
which++;
}
FClose(intree);
if (numtrees > 1 && loci > 1 ) {
weight = (long *)Malloc(loci*sizeof(long));
for (i = 0; i < loci; i++)
weight[i] = 1;
standev2(numtrees, maxwhich, 0, loci-1, maxlogl, l0gl, l0gf, seed);
free(weight);
fprintf(outfile, "\n\n");
}
} else { /* if ( !usertree ) */
if (jumb == 1) {
setuptree(&curtree, nonodes2);
setuptree(&priortree, nonodes2);
setuptree(&bestree, nonodes2);
if (njumble > 1)
setuptree(&bestree2, nonodes2);
}
for (i = 1; i <= spp; i++)
enterorder[i - 1] = i;
if (jumble)
randumize(seed, enterorder);
nextsp = 3;
buildsimpletree(&curtree);
curtree.start = curtree.nodep[enterorder[0] - 1]->back;
if (jumb == 1) numtrees = 1;
nextsp = 4;
if (progress) {
printf("Adding species:\n");
writename(0, 3, enterorder);
#ifdef WIN32
phyFillScreenColor();
#endif
}
while (nextsp <= spp) {
buildnewtip(enterorder[nextsp - 1], &curtree, nextsp);
copy_(&curtree, &priortree);
bestree.likelihood = -DBL_MAX;
addtraverse(curtree.nodep[enterorder[nextsp - 1] - 1]->back,
curtree.start, true );
copy_(&bestree, &curtree);
if (progress) {
writename(nextsp - 1, 1, enterorder);
#ifdef WIN32
phyFillScreenColor();
#endif
}
if (global && nextsp == spp) {
if (progress) {
printf("\nDoing global rearrangements\n");
printf(" !");
for (i = 1; i <= spp - 2; i++)
if ( (i - spp) % (( nonodes2 / 72 ) + 1 ) == 0 )
putchar('-');
printf("!\n");
printf(" ");
}
}
succeeded = true;
while (succeeded) {
succeeded = false;
if ( global && nextsp == spp )
globrearrange();
else
rearrange(curtree.start);
if (global && nextsp == spp)
putc('\n', outfile);
}
if (global && nextsp == spp && progress)
putchar('\n');
if (njumble > 1) {
if (jumb == 1 && nextsp == spp)
copy_(&bestree, &bestree2);
else if (nextsp == spp) {
if (bestree2.likelihood < bestree.likelihood)
copy_(&bestree, &bestree2);
}
}
if (nextsp == spp && jumb == njumble) {
if (njumble > 1) copy_(&bestree2, &curtree);
curtree.start = curtree.nodep[outgrno - 1]->back;
printree();
summarize();
}
nextsp++;
}
}
if ( jumb < njumble)
return;
if (progress) {
printf("\nOutput written to file \"%s\"\n", outfilename);
if (trout)
printf("\nTree also written onto file \"%s\"\n\n", outtreename);
}
freeview(&curtree, nonodes2);
if (!usertree) {
freeview(&bestree, nonodes2);
freeview(&priortree, nonodes2);
}
for (i = 0; i < spp; i++)
free(x[i]);
if (!contchars) {
free(locus);
free(pbar);
}
} /* maketree */
ajuint embMatProtScanInt(const AjPStr s, const AjPStr n, const EmbPMatPrints m,
AjPList *l,
AjBool *all, AjBool *ordered, AjBool overlap)
{
EmbPMatMatch mm;
AjPStr t;
char *p;
char *q;
ajint slen;
ajint score;
ajint mlen;
ajint elem;
ajint minpc;
ajint maxscore;
ajint limit;
ajint sum;
ajint hpe;
ajint hpm;
ajint lastelem;
ajint lastpos;
ajint op;
ajint i;
ajint j;
t = ajStrNewC(ajStrGetPtr(s));
ajStrFmtUpper(&t);
p = q = ajStrGetuniquePtr(&t);
slen = ajStrGetLen(t);
for(i=0;i<slen;++i,++p)
*p = ajSysCastItoc(ajBasecodeToInt((ajint)*p));
p = q;
*all = *ordered = ajTrue;
lastelem = lastpos = INT_MAX;
hpm=0;
for(elem=(m)->n - 1;elem >= 0;--elem)
{
hpe = 0;
mlen = (m)->len[elem];
minpc = (m)->thresh[elem];
maxscore = (m)->max[elem];
limit = slen-mlen;
for(i=0;i<limit;++i)
{
sum = 0;
for(j=0;j<mlen;++j)
sum += (m)->matrix[elem][(ajint) p[i+j]][j];
score = (sum*100)/maxscore;
if(score>=minpc)
{
if(elem<lastelem && *ordered)
{
if(lastelem == INT_MAX)
{
lastelem = elem;
lastpos = i;
}
else
{
lastelem = elem;
op = i;
if(!overlap)
op += mlen;
if(op >= lastpos)
*ordered = ajFalse;
lastpos = i;
}
}
++hpe;
++hpm;
matPushHitInt(n,m,l,i,score,elem,hpe,hpm);
}
}
if(!hpe)
*all = ajFalse;
}
if(hpm)
{
ajListPop(*l,(void **)&mm);
if(*all)
{
mm->all = ajTrue;
if(*ordered)
mm->ordered = ajTrue;
else
mm->ordered = ajFalse;
}
else
{
mm->all = ajFalse;
if(*ordered)
mm->ordered = ajTrue;
else
mm->ordered = ajFalse;
}
ajListPush(*l,(void *)mm);
}
ajStrDel(&t);
return hpm;
}
static PSilent silent_checktrans(const AjPStr seq,const EmbPMatMatch match,
const PRinfo rlp, ajint begin, ajint radj,
AjBool rev, ajint end)
{
PSilent ret;
const char *p = NULL;
const char *q = NULL;
const char *s = NULL;
char *t;
const char *u;
ajint matchpos;
ajint framep;
ajint count;
AjPTrn table = NULL;
AjPStr s1 = NULL;
AjPStr s2 = NULL;
char c;
char rc;
ajint min = INT_MAX; /* Reverse sense intentional! */
ajint max = -INT_MAX;
ajint fpos;
ajint rpos;
ajint x;
AjPStr tstr = NULL;
matchpos = match->start;
fpos = matchpos;
rpos=radj-fpos-match->len;
tstr = ajStrNewS(seq);
t = ajStrGetuniquePtr(&tstr);
p = t+fpos-(begin+1);
u = q = ajStrGetPtr(rlp->site);
/* Test here for whether cut site is within sequence substring */
if(rlp->ncuts==4)
{
min = AJMIN(rlp->cut1,rlp->cut2);
max = AJMAX(rlp->cut3,rlp->cut4);
}
else if(rlp->ncuts==2)
{
min = AJMIN(rlp->cut1,rlp->cut2);
max = AJMAX(rlp->cut1,rlp->cut2);
}
else
{
ajWarn("Possibly corrupt RE file");
ajStrDel(&tstr);
return NULL;
}
if(!rev) /* forward strand */
{
if(matchpos+min<0||matchpos+max>end+1)
{
/*Cut site not in sequence range*/
ajStrDel(&tstr);
return NULL;
}
}
else /* reverse strand */
{
if(radj-matchpos-1-min>end+1||radj-matchpos-1-max<begin)
{
/*Cut site not in sequence range*/
ajStrDel(&tstr);
return NULL;
}
}
count=0;
while(ajBaseAlphaToBin(*q++) & ajBaseAlphaToBin(*p++))
++count;
/* Changed base postion */
x = fpos+count-(begin+1);
/* Where the frame starts on the reverse strand */
framep = (end-begin+1)%3;
c = t[x];
rc = u[count];
if(!rev) /* forward strand */
s = t+x-x%3;
else /* reverse strand */
s = t+x-(x-framep)%3;
table = ajTrnNewI(0);
/* translates codon pointed to by s (original seq) */
s1 = ajStrNewK(ajTrnCodonC(table,s));
t[x] = rc;
/* translates codon pointed to by s (mutated base from RS pattern */
s2 = ajStrNewK(ajTrnCodonC(table,s));
t[x] = c; /* changes mutated base in seq back to original base */
AJNEW(ret);
ret->obase = c;
ret->nbase = rc;
ret->code = ajStrNewC(ajStrGetPtr(rlp->code));
ret->site = ajStrNewC(ajStrGetPtr(rlp->site));
ret->seqaa = ajStrNewC(ajStrGetPtr(s1));
ret->reaa = ajStrNewC(ajStrGetPtr(s2));
if(ajStrMatchS(s1,s2))
ret->issilent = ajTrue;
else
ret->issilent = ajFalse;
if(!rev)
{
ret->match = matchpos;
ret->base = matchpos+count;
}
else
{
ret->match = rpos;
ret->base = rpos+match->len-1-count;
}
ajStrDel(&tstr);
ajStrDel(&s1);
ajStrDel(&s2);
ajTrnDel(&table);
return ret;
}
static char* cutgextract_next(AjPFile inf, const AjPStr wildspecies,
AjPStr* pspecies, AjPStr* pdoc)
{
AjPStrTok handle = NULL;
AjPStr token = NULL;
ajint i;
ajint len;
char *p = NULL;
char c;
AjBool done = ajFalse;
if(!cutgextractLine)
cutgextractLine = ajStrNew();
if(!cutgextractOrg)
cutgextractOrg = ajStrNew();
ajStrAssignC(&cutgextractLine,"");
ajStrAssignC(pdoc,"");
while (!done)
{
while(ajStrGetCharFirst(cutgextractLine) != '>')
if(!ajReadlineTrim(inf,&cutgextractLine))
return NULL;
handle = ajStrTokenNewC(cutgextractLine,"\\\n\t\r");
for(i=0;i<7;++i) {
ajStrTokenNextParseC(&handle,"\\\n\t\r",&token);
if(i==5)
{
ajStrAssignC(&cutgextractOrg,"E");
ajStrAppendS(&cutgextractOrg, token);
ajStrAssignS(pspecies, token);
if(ajStrMatchWildS(token,wildspecies))
{
done = ajTrue;
}
}
switch(i)
{
case 0:
ajStrAppendC(pdoc, "#ID ");
ajStrAppendS(pdoc, token);
ajStrAppendC(pdoc, "\n");
break;
case 1:
ajStrAppendC(pdoc, "#AC ");
ajStrAppendS(pdoc, token);
ajStrAppendC(pdoc, "\n");
break;
case 2:
ajStrAppendC(pdoc, "#FT ");
ajStrAppendS(pdoc, token);
ajStrAppendC(pdoc, "\n");
break;
case 3:
ajStrAppendC(pdoc, "#FL ");
ajStrAppendS(pdoc, token);
ajStrAppendC(pdoc, "\n");
break;
case 4:
ajStrAppendC(pdoc, "#PI ");
ajStrAppendS(pdoc, token);
ajStrAppendC(pdoc, "\n");
ajStrAssignS(&cutgextractSavepid, token);
break;
case 5:
ajStrAppendC(pdoc, "#OS ");
ajStrAppendS(pdoc, token);
ajStrAppendC(pdoc, "\n");
break;
case 6:
ajStrAppendC(pdoc, "#DE ");
ajStrAppendS(pdoc, token);
ajStrAppendC(pdoc, "\n");
break;
default:
break;
}
}
ajStrTokenDel(&handle);
ajStrDel(&token);
if(!done)
if(!ajReadlineTrim(inf,&cutgextractLine))
return NULL;
}
len = ajStrGetLen(cutgextractOrg);
p = ajStrGetuniquePtr(&cutgextractOrg);
for(i=0;i<len;++i)
{
c = p[i];
if(c=='/' || c==' ' || c=='.' || c=='\'')
p[i]='_';
}
if(p[strlen(p)-1]=='_')
p[strlen(p)-1]='\0';
ajStrDel(&token);
ajStrTokenDel(&handle);
return p;
}
int main(int argc, Char *argv[])
{
pattern_elm ***pattern_array;
long tip_count = 0;
double ln_maxgrp;
double ln_maxgrp1;
double ln_maxgrp2;
node * p;
#ifdef MAC
argc = 1; /* macsetup("Treedist", ""); */
argv[0] = "Treedist";
#endif
init(argc, argv);
emboss_getoptions("ftreedist",argc,argv);
/* Initialize option-based variables, then ask for changes regarding
their values. */
ntrees = 0.0;
lasti = -1;
/* read files to determine size of structures we'll be working with */
countcomma(ajStrGetuniquePtr(&phylotrees[0]->Tree),&tip_count);
tip_count++; /* countcomma does a raw comma count, tips is one greater */
/*
* EWFIX.BUG.756 -- this section may be killed if a good solution
* to bug 756 is found
*
* inside cons.c there are several arrays which are allocated
* to size "maxgrp", the maximum number of groups (sets of
* tips more closely connected than the rest of the tree) we
* can see as the code executes.
*
* We have two measures we use to determine how much space to
* allot:
* (1) based on the tip count of the trees in the infile
* (2) based on total number of trees in infile, and
*
* (1) -- Tip Count Method
* Since each group is a subset of the set of tips we must
* represent at most pow(2,tips) different groups. (Technically
* two fewer since we don't store the empty or complete subsets,
* but let's keep this simple.
*
* (2) -- Total Tree Size Method
* Each tree we read results in
* singleton groups for each tip, plus
* a group for each interior node except the root
* Since the singleton tips are identical for each tree, this gives
* a bound of #tips + ( #trees * (# tips - 2 ) )
*
*
* Ignoring small terms where expedient, either of the following should
* result in an adequate allocation:
* pow(2,#tips)
* (#trees + 1) * #tips
*
* Since "maxgrp" is a limit on the number of items we'll need to put
* in a hash, we double it to make space for quick hashing
*
* BUT -- all of this has the possibility for overflow, so -- let's
* make the initial calculations with doubles and then convert
*
*/
/* limit chosen to make hash arithmetic work */
maxgrp = LONG_MAX / 2;
ln_maxgrp = log((double)maxgrp);
/* 2 * (#trees + 1) * #tips */
ln_maxgrp1 = log(2.0 * (double)tip_count *
((double)trees_in_1 + (double)trees_in_2));
/* ln only for 2 * pow(2,#tips) */
ln_maxgrp2 = (double)(1 + tip_count) * log(2.0);
/* now -- find the smallest of the three */
if(ln_maxgrp1 < ln_maxgrp)
{
maxgrp = 2 * (trees_in_1 + trees_in_2 + 1) * tip_count;
ln_maxgrp = ln_maxgrp1;
}
if(ln_maxgrp2 < ln_maxgrp)
{
maxgrp = pow(2,tip_count+1);
}
/* Read the (first) tree file and put together grouping, order, and
timesseen */
read_groups (&pattern_array, trees_in_1 + trees_in_2, tip_count, phylotrees);
if ((tree_pairing == ADJACENT_PAIRS) ||
(tree_pairing == ALL_IN_FIRST)) {
/* Here deal with the adjacent or all-in-first pairing
difference computation */
compute_distances (pattern_array, trees_in_1, 0);
} else if (tree_pairing == NO_PAIRING) {
/* Compute the consensus tree. */
putc('\n', outfile);
/* consensus(); Reserved for future development */
}
if (progress)
printf("\nOutput written to file \"%s\"\n\n", outfilename);
FClose(outtree);
FClose(intree);
FClose(outfile);
if ((tree_pairing == ALL_IN_1_AND_2) ||
(tree_pairing == CORR_IN_1_AND_2))
FClose(intree2);
#ifdef MAC
fixmacfile(outfilename);
fixmacfile(outtreename);
#endif
free_patterns (pattern_array, trees_in_1 + trees_in_2);
clean_up_final();
/* clean up grbg */
p = grbg;
while (p != NULL) {
node * r = p;
p = p->next;
free(r->nodeset);
free(r->view);
free(r);
}
printf("Done.\n\n");
embExit();
return 0;
} /* main */
