/* @funcstatic domainreso_StrBinSearchDomain ********************************
**
** Performs a binary search for a DOMAIN domain code over an array of AjPStr
** (which of course must first have been sorted, e.g. by domainreso_StrComp).
** This is a case-insensitive search.
**
** @param [r] id [AjPStr] Search term
** @param [r] arr [AjPStr*] Array of AjPStr objects
** @param [r] siz [ajint] Size of array
**
** @return [ajint] Index of first AjPStr object found with an PDB code
** matching id, or -1 if id is not found.
** @@
******************************************************************************/
static ajint domainreso_StrBinSearchDomain(AjPStr id,
AjPStr *arr,
ajint siz)
{
int l;
int m;
int h;
int c;
l = 0;
h = siz-1;
while(l<=h)
{
m = (l+h)>>1;
if((c=ajStrCmpCaseS(id, arr[m])) < 0)
h = m-1;
else if(c>0)
l = m+1;
else
return m;
}
return -1;
}
static int cacheensembl_stringcompare(const void* P1, const void* P2)
{
const AjPStr string1 = NULL;
const AjPStr string2 = NULL;
string1 = *(AjPStr const*) P1;
string2 = *(AjPStr const*) P2;
#if AJFALSE
if(ajDebugTest("cacheensembl_stringcompare"))
ajDebug("cacheensembl_stringcompare\n"
" string1 %u\n"
" string2 %u\n",
string1,
string2);
#endif
return ajStrCmpCaseS(string1, string2);
}
static void remap_NoCutList(AjPFile outfile, const AjPTable hittable,
AjBool html, const AjPStr enzymes, AjBool blunt,
AjBool sticky, ajuint sitelen, AjBool commercial,
AjBool ambiguity, AjBool limit,
const AjPTable retable)
{
/* for iterating over hittable */
PValue value;
void **keyarray = NULL; /* array for table */
void **valarray = NULL; /* array for table */
ajint i;
/* list of enzymes that cut */
AjPList cutlist;
AjIList citer; /* iterator for cutlist */
AjPStr cutname = NULL;
AjBool found;
/* for parsing value->iso string */
AjPStrTok tok;
char tokens[] = " ,";
AjPStr code = NULL;
const char *p;
/* for reading in enzymes names */
AjPFile enzfile = NULL;
AjPStr *ea;
ajint ne; /* number of enzymes */
AjBool isall = ajTrue;
/* list of enzymes that don't cut */
AjPList nocutlist;
AjIList niter; /* iterator for nocutlist */
AjPStr nocutname = NULL;
/* count of rejected enzymes not matching criteria */
ajint rejected_count = 0;
EmbPPatRestrict enz;
/* for renaming preferred isoschizomers */
AjPList newlist;
/*
**
** Make a list of enzymes('cutlist') that hit
** including the isoschizomer names
**
*/
ajDebug("Make a list of all enzymes that cut\n");
cutlist = ajListstrNew();
nocutlist = ajListstrNew();
ajTableToarrayKeysValues(hittable, &keyarray, &valarray);
for(i = 0; keyarray[i]; i++)
{
value = (PValue) valarray[i];
cutname = ajStrNew();
ajStrAssignRef(&cutname, keyarray[i]);
ajListstrPushAppend(cutlist, cutname);
/* Add to cutlist all isoschizomers of enzymes that cut */
ajDebug("Add to cutlist all isoschizomers of enzymes that cut\n");
/* start token to parse isoschizomers names */
tok = ajStrTokenNewC(value->iso, tokens);
while(ajStrTokenNextParseC(&tok, tokens, &code))
{
cutname = ajStrNew();
ajStrAssignS(&cutname, code);
ajListstrPushAppend(cutlist, cutname);
}
ajStrTokenDel(&tok);
}
ajStrDel(&code);
AJFREE(keyarray);
AJFREE(valarray);
/*
** Read in list of enzymes ('nocutlist') - either all or
** the input enzyme list.
** Exclude those that don't match the selection criteria - count these.
*/
ajDebug("Read in a list of all input enzyme names\n");
ne = 0;
if(!enzymes)
isall = ajTrue;
else
{
/* get input list of enzymes into ea[] */
ne = ajArrCommaList(enzymes, &ea);
if(ajStrMatchCaseC(ea[0], "all"))
isall = ajTrue;
else
{
isall = ajFalse;
for(i=0; i<ne; ++i)
ajStrRemoveWhite(&ea[i]);
}
}
enzfile = ajDatafileNewInNameC(ENZDATA);
/* push all enzyme names without the required criteria onto nocutlist */
enz = embPatRestrictNew();
while(!ajFileIsEof(enzfile))
{
if(!embPatRestrictReadEntry(enz, enzfile))
continue;
/*
** If user entered explicit enzyme list, then check to see if
** this is one of that explicit list
*/
if(!isall)
{
found = AJFALSE;
for(i=0; i<ne; ++i)
if(ajStrMatchCaseS(ea[i], enz->cod))
{
found = AJTRUE;
break;
}
if(!found) /* not in the explicit list */
continue;
ajDebug("RE %S is in the input explicit list of REs\n", enz->cod);
}
/* ignore ncuts==0 as they are unknown */
if(!enz->ncuts)
{
/* number of cut positions */
ajDebug("RE %S has an unknown number of cut positions\n",
enz->cod);
continue;
}
ajDebug("RE %S has a known number of cut sites\n", enz->cod);
if(enz->len < sitelen)
{
/* recognition site length */
ajDebug("RE %S does not have a long enough recognition site\n",
enz->cod);
rejected_count++;
continue;
}
if(!blunt && enz->blunt)
{
/* blunt/sticky */
ajDebug("RE %S is blunt\n", enz->cod);
rejected_count++;
continue;
}
if(!sticky && !enz->blunt)
{
/* blunt/sticky */
ajDebug("RE %S is sticky\n", enz->cod);
rejected_count++;
continue;
}
/* commercially available enzymes have uppercase patterns */
p = ajStrGetPtr(enz->pat);
/*
** The -commercial qualifier is only used if we are searching
** through 'all' of the REBASE database - if we have specified an
** explicit list of enzymes then they are searched for whether or
** not they are commercially available
*/
if((*p >= 'a' && *p <= 'z') && commercial && isall)
{
ajDebug("RE %S is not commercial\n", enz->cod);
rejected_count++;
continue;
}
if(!ambiguity && remap_Ambiguous(enz->pat)) {
ajDebug("RE %S is ambiguous\n", enz->cod);
rejected_count++;
continue;
}
ajDebug("RE %S matches all required criteria\n", enz->cod);
code = ajStrNew();
ajStrAssignS(&code, enz->cod);
ajListstrPushAppend(nocutlist, code);
}
embPatRestrictDel(&enz);
ajFileClose(&enzfile);
for(i=0; i<ne; ++i)
if(ea[i])
ajStrDel(&ea[i]);
if(ne)
AJFREE(ea);
/*
** Change names of enzymes in the non-cutter list
** to that of preferred (prototype)
** enzyme name so that the isoschizomers of cutters
** will be removed from the
** non-cutter list in the next bit.
** Remove duplicate prototype names.
*/
if(limit)
{
newlist = ajListstrNew();
remap_RenamePreferred(nocutlist, retable, newlist);
ajListstrFreeData(&nocutlist);
nocutlist = newlist;
ajListSortUnique(nocutlist, remap_cmpcase, remap_strdel);
}
/*
** Iterate through the list of input enzymes removing those that are in
** the cutlist.
*/
ajDebug("Remove from the nocutlist all enzymes and isoschizomers "
"that cut\n");
/*
** This steps down both lists at the same time, comparing names and
** iterating to the next name in whichever list whose name compares
** alphabetically before the other. Where a match is found, the
** nocutlist item is deleted.
*/
ajListSort(nocutlist, remap_cmpcase);
ajListSort(cutlist, remap_cmpcase);
citer = ajListIterNewread(cutlist);
niter = ajListIterNew(nocutlist);
/*
while((cutname = (AjPStr)ajListIterGet(citer)) != NULL)
ajDebug("dbg cutname = %S\n", cutname);
*/
nocutname = (AjPStr)ajListIterGet(niter);
cutname = (AjPStr)ajListIterGet(citer);
ajDebug("initial cutname, nocutname: '%S' '%S'\n", cutname, nocutname);
while(nocutname != NULL && cutname != NULL)
{
i = ajStrCmpCaseS(cutname, nocutname);
ajDebug("compare cutname, nocutname: %S %S ", cutname, nocutname);
ajDebug("ajStrCmpCase=%d\n", i);
if(i == 0)
{
/* match - so remove from nocutlist */
ajDebug("ajListstrRemove %S\n", nocutname);
ajListstrIterRemove(niter);
nocutname = (AjPStr)ajListIterGet(niter);
/*
** Don't increment the cutname list pointer here
** - there may be more than one entry in the nocutname
** list with the same name because we have converted
** isoschizomers to their preferred name
*/
/* cutname = (AjPStr)ajListIterGet(citer); */
}
else if(i == -1)
/* cutlist name sorts before nocutlist name */
cutname = (AjPStr)ajListIterGet(citer);
else if(i == 1)
/* nocutlist name sorts before cutlist name */
nocutname = (AjPStr)ajListIterGet(niter);
}
ajListIterDel(&citer);
ajListIterDel(&niter);
ajListstrFreeData(&cutlist);
/* Print the resulting list of those that do not cut*/
ajDebug("Print out the list\n");
/* print the title */
if(html)
ajFmtPrintF(outfile, "<H2>");
ajFmtPrintF(outfile, "\n\n# Enzymes that do not cut\n\n");
if(html)
ajFmtPrintF(outfile, "</H2>\n");
if(html)
ajFmtPrintF(outfile, "<PRE>");
/* ajListSort(nocutlist, ajStrVcmp);*/
niter = ajListIterNewread(nocutlist);
i = 0;
while((nocutname = (AjPStr)ajListIterGet(niter)) != NULL)
{
ajFmtPrintF(outfile, "%-10S", nocutname);
/* new line after every 7 names printed */
if(i++ == 7)
{
ajFmtPrintF(outfile, "\n");
i = 0;
}
}
ajListIterDel(&niter);
/* end the output */
ajFmtPrintF(outfile, "\n");
if(html) {ajFmtPrintF(outfile, "</PRE>\n");}
/*
** Print the count of rejected enzymes
** N.B. This is the count of ALL rejected enzymes including all
** isoschizomers
*/
if(html)
ajFmtPrintF(outfile, "<H2>");
ajFmtPrintF(outfile,
"\n\n# No. of cutting enzymes which do not match the\n"
"# SITELEN, BLUNT, STICKY, COMMERCIAL, AMBIGUOUS criteria\n\n");
if(html)
ajFmtPrintF(outfile, "</H2>\n");
ajFmtPrintF(outfile, "%d\n", rejected_count);
ajDebug("Tidy up\n");
ajListstrFreeData(&nocutlist);
ajListstrFreeData(&cutlist);
return;
}
static void remap_RestrictPreferred(const AjPList l, const AjPTable t)
{
AjIList iter = NULL;
EmbPMatMatch m = NULL;
const AjPStr value = NULL;
AjPStr newiso = NULL;
AjBool found; /* name found in isoschizomer list */
/* for parsing value->iso string */
AjPStrTok tok = NULL;
char tokens[] = " ,";
AjPStr code = NULL;
iter = ajListIterNewread(l);
while((m = (EmbPMatMatch)ajListIterGet(iter)))
{
found = ajFalse;
/* get prototype name */
value = ajTableFetchS(t, m->cod);
if(value)
{
ajStrAssignC(&newiso, "");
/* parse isoschizomer names from m->iso */
ajStrTokenDel(&tok);
tok = ajStrTokenNewC(m->iso, tokens);
while(ajStrTokenNextParseC(&tok, tokens, &code))
{
if(ajStrGetLen(newiso) > 0)
ajStrAppendC(&newiso, ",");
/* found the prototype name? */
if(!ajStrCmpCaseS(code, value))
{
ajStrAppendS(&newiso, m->cod);
found = ajTrue;
}
else
ajStrAppendS(&newiso, code);
}
ajStrTokenDel(&tok);
/* if the name was not replaced, then add it in now */
if(!found)
{
if(ajStrGetLen(newiso) > 0)
ajStrAppendC(&newiso, ",");
ajStrAppendS(&newiso, m->cod);
}
ajDebug("RE: %S -> %S iso=%S newiso=%S\n", m->cod, value,
m->iso, newiso);
/* replace the old iso string with the new one */
ajStrAssignS(&m->iso, newiso);
/* rename the enzyme to the prototype name */
ajStrAssignS(&m->cod, value);
}
}
ajListIterDel(&iter);
ajStrDel(&newiso);
ajStrDel(&code);
ajStrTokenDel(&tok);
return;
}
int main(int argc, char **argv)
{
AjPSeqset seq1;
AjPSeqset seq2;
AjPFile list;
ajint n1;
ajint n2;
ajint *lengths1;
ajint *lengths2;
ajuint *order1;
ajuint *order2;
ajint *hits1;
ajint *hits2;
ajint curr1;
ajint curr2;
ajint tmp1;
ajint tmp2 = 0;
ajint i;
AjPStr operator;
ajint OperatorCode=0;
embInit("listor", argc, argv);
seq1 = ajAcdGetSeqset("firstsequences");
seq2 = ajAcdGetSeqset("secondsequences");
list = ajAcdGetOutfile("outfile");
operator = ajAcdGetListSingle("operator");
/* get the operator value */
switch(ajStrGetCharFirst(operator))
{
case 'O':
OperatorCode = L_OR;
break;
case 'A':
OperatorCode = L_AND;
break;
case 'X':
OperatorCode = L_XOR;
break;
case 'N':
OperatorCode = L_NOT;
break;
default:
ajFatal("Invalid operator type: %S", operator);
embExitBad();
}
/* get the order of seqset 1 by length */
n1 = ajSeqsetGetSize(seq1);
/* lengths of seq1 entries */
lengths1 = AJCALLOC0(n1, sizeof(ajint));
/* seq1 entries which match seq2 */
hits1 = AJCALLOC0(n1, sizeof(ajint));
/* seq1 entries in length order */
order1 = AJCALLOC0(n1, sizeof(ajint));
for(i=0; i<n1; i++)
{
lengths1[i] = ajSeqGetLen(ajSeqsetGetseqSeq(seq1, i));
order1[i] = i;
hits1[i] = -1;
}
ajSortIntIncI(lengths1, order1, n1);
/* get the order of seqset 2 by length */
n2 = ajSeqsetGetSize(seq2);
lengths2 = AJCALLOC0(n2, sizeof(ajint));
hits2 = AJCALLOC0(n2, sizeof(ajint));
order2 = AJCALLOC0(n2, sizeof(ajint));
for(i=0; i<n2; i++)
{
lengths2[i] = ajSeqGetLen(ajSeqsetGetseqSeq(seq2, i));
order2[i] = i;
hits2[i] = -1;
}
ajSortIntIncI(lengths2, order2, n2);
/*
** go down the two sequence sets, by size order, looking for identical
**lengths
*/
curr1 = 0;
curr2 = 0;
while(curr1 < n1 && curr2 < n2)
{
if(lengths1[order1[curr1]] < lengths2[order2[curr2]])
/* seq1 is shorter - increment curr1 index */
curr1++;
else if(lengths1[order1[curr1]] > lengths2[order2[curr2]])
/* seq2 is shorter - increment curr2 index */
curr2++;
else
{
/* identical lengths - check all seq1/seq2 entries of this len */
for(tmp1=curr1; tmp1<n1
&& lengths1[order1[tmp1]] == lengths2[order2[curr2]]; tmp1++)
for(tmp2=curr2; tmp2<n2 && lengths2[order2[tmp2]] ==
lengths2[order2[curr2]]; tmp2++)
/* check to see if the sequences are identical */
if(!ajStrCmpCaseS(ajSeqGetSeqS(ajSeqsetGetseqSeq(seq1,
order1[tmp1])),
ajSeqGetSeqS(ajSeqsetGetseqSeq(seq2,
order2[tmp2]))))
{
hits1[order1[tmp1]] = order2[tmp2];
hits2[order2[tmp2]] = order1[tmp1];
}
curr1 = tmp1;
curr2 = tmp2;
}
}
/* output the required entries to the list file */
listor_Output(list, OperatorCode, seq1, seq2, hits1, hits2, n1, n2);
AJFREE(lengths1);
AJFREE(lengths2);
AJFREE(order1);
AJFREE(order2);
AJFREE(hits1);
AJFREE(hits2);
ajFileClose(&list);
ajStrDel(&operator);
ajSeqsetDel(&seq1);
ajSeqsetDel(&seq2);
embExit();
return 0;
}
