/* @funcstatic seqwords_keysearch ********************************************
**
** Search swissprot with terms structure and writes a hitlist structure
**
** @param [r] inf [AjPFile] File pointer to swissprot database
** @param [r] terms [AjPTerms] Terms object pointer
** @param [w] hits [EmbPHitlist*] Hitlist object pointer
**
** @return [AjBool] True on success
** @@
******************************************************************************/
static AjBool seqwords_keysearch(AjPFile inf,
AjPTerms terms,
EmbPHitlist *hits)
{
AjPStr line =NULL; /* Line of text. */
AjPStr id =NULL; /* Line of text. */
AjPStr temp =NULL;
ajint s =0; /* Temp. start of hit value. */
ajint e =0; /* Temp. end of hit value. */
AjPInt start =NULL; /* Array of start of hit(s). */
AjPInt end =NULL; /* Array of end of hit(s). */
ajint nhits =0; /* Number of hits. */
ajint x =0;
AjBool foundkw =ajFalse;
AjBool foundft =ajFalse;
/* Check for valid args. */
if(!inf)
return ajFalse;
/* Allocate strings and arrays. */
line = ajStrNew();
id = ajStrNew();
start = ajIntNew();
end = ajIntNew();
/* Start of main loop. */
while((ajReadlineTrim(inf,&line)))
{
/* Parse the AC line. */
if(ajStrPrefixC(line,"AC"))
{
/* Copy accesion number and remove the ';' from the end. */
ajFmtScanS(line, "%*s %S", &id);
ajStrExchangeCC(&id, ";", "\0");
/* Reset flags & no. hits. */
foundkw=ajFalse;
foundft=ajFalse;
nhits=0;
}
/* Search the description and keyword lines with search terms. */
else if((ajStrPrefixC(line,"DE") || (ajStrPrefixC(line,"KW"))))
{
/*
** Search terms have a leading and trailing space to prevent
** them being found as substrings within other words. To
** catch cases where a DE or KW line begins with a search
** term, we must add a leading and trailing space to line.
** We must first remove punctation from the line to be parsed.
*/
ajStrExchangeSetCC(&line, ".,;:", " ");
ajStrAppendK(&line, ' ');
ajStrInsertC(&line, 0, " ");
for (x = 0; x < terms->N; x++)
/* Search term is found. */
if((ajStrFindCaseS(line, terms->Keywords[x])!=-1))
{
foundkw=ajTrue;
break;
}
}
/* Search the feature table line with search terms. */
else if((ajStrPrefixC(line,"FT DOMAIN")))
{
/*
** Search terms have a leading and trailing space to prevent
** them being found as substrings within other words. To
** catch cases where a FT line ends with a search
** term, we must add a trailing space to line
** We must first remove punctation from the line to be parsed.
*/
ajStrExchangeSetCC(&line, ".,;:", " ");
ajStrAppendK(&line, ' ');
for (x = 0; x < terms->N; x++)
if((ajStrFindCaseS(line, terms->Keywords[x])!=-1))
{
/* Search term is found. */
foundft = ajTrue;
nhits++;
/* Assign start and end of hit. */
ajFmtScanS(line, "%*s %*s %d %d", &s, &e);
ajIntPut(&start, nhits-1, s);
ajIntPut(&end, nhits-1, e);
break;
}
}
/* Parse the sequence. */
else if((ajStrPrefixC(line,"SQ") && ((foundkw == ajTrue) ||
(foundft == ajTrue))))
{
/* Allocate memory for temp. sequence. */
temp = ajStrNew();
/* Read the sequence into hitlist structure. */
while((ajReadlineTrim(inf,&line)) && !ajStrPrefixC(line,"//"))
/* Read sequence line into temp. */
ajStrAppendC(&temp,ajStrGetPtr(line)+3);
/* Clean up temp. sequence. */
ajStrRemoveWhite(&temp);
/*Priority is given to domain (rather than full length) sequence.*/
if(foundft)
{
for(x=0;x<nhits;x++)
{
/* Increment counter of hits for subsequent hits*/
(*hits)->N++;
/* Reallocate memory for array of hits in hitlist
structure. */
AJCRESIZE((*hits)->hits, (*hits)->N);
(*hits)->hits[(*hits)->N-1]=embHitNew();
ajStrAssignC(&(*hits)->hits[(*hits)->N-1]->Model,
"KEYWORD");
/* Assign start and end of hit. */
(*hits)->hits[(*hits)->N-1]->Start = ajIntGet(start, x);
(*hits)->hits[(*hits)->N-1]->End = ajIntGet(end, x);
/* Extract sequence within specified range */
ajStrAssignSubS(&(*hits)->hits[(*hits)->N - 1]->Seq, temp,
(*hits)->hits[(*hits)->N - 1]->Start - 1,
(*hits)->hits[(*hits)->N - 1]->End - 1);
/* Put id into structure */
ajStrAssignRef(&(*hits)->hits[(*hits)->N - 1]->Acc, id);
}
}
else
{
/* Increment counter of hits */
(*hits)->N++;
/* Reallocate memory for array of hits in hitlist structure */
AJCRESIZE((*hits)->hits, (*hits)->N);
(*hits)->hits[(*hits)->N-1]=embHitNew();
ajStrAssignC(&(*hits)->hits[(*hits)->N-1]->Model, "KEYWORD");
/* Extract whole sequence */
ajStrAssignRef(&(*hits)->hits[(*hits)->N - 1]->Seq, temp);
(*hits)->hits[(*hits)->N - 1]->Start = 1;
(*hits)->hits[(*hits)->N - 1]->End =
ajStrGetLen((*hits)->hits[(*hits)->N - 1]->Seq);
/* Put id into structure */
ajStrAssignRef(&(*hits)->hits[(*hits)->N - 1]->Acc, id);
}
/* Free temp. sequence */
ajStrDel(&temp);
}
}
/* Clean up */
ajStrDel(&line);
ajStrDel(&id);
ajIntDel(&start);
ajIntDel(&end);
return ajTrue;
}
static void domainalign_ProcessStampFile(AjPStr in,
AjPStr out,
AjPDomain domain,
ajint noden,
AjPFile logf)
{
AjPFile outf = NULL; /* Output file pointer. */
AjPFile inf = NULL; /* Input file pointer. */
AjPStr temp1 = NULL; /* Temporary string. */
AjPStr temp2 = NULL; /* Temporary string. */
AjPStr temp3 = NULL; /* Temporary string. */
AjPStr line = NULL; /* Line of text from input file. */
ajint blk = 1; /* Count of the current block in the input file.
Block 1 is the numbering and protein sequences,
Block 2 is the secondary structure,
Block 3 is the Very/Less/Post similar records*/
AjBool ok = ajFalse;
/* Initialise strings. */
line = ajStrNew();
temp1 = ajStrNew();
temp2 = ajStrNew();
temp3 = ajStrNew();
/* Open input and output files. */
if(!(inf=ajFileNewInNameS(in)))
ajFatal("Could not open input file in domainalign_ProcessStampFile");
/* Start of code for reading input file.
Ignore everything up to first line beginning with 'Number'. */
while((ajReadlineTrim(inf,&line)))
{
/* ajFileReadLine will trim the tailing \n. */
if((ajStrGetCharPos(line, 1)=='\0'))
{
ok = ajTrue;
break;
}
}
/* Read rest of input file. */
if(ok)
{
/* Write DOMAIN classification records to file. */
if(!(outf=ajFileNewOutNameS(out)))
ajFatal("Could not open output file in domainalign_ProcessStampFile");
if((domain->Type == ajSCOP))
{
ajFmtPrintF(outf,"# TY SCOP\n# XX\n");
ajFmtPrintF(outf,"# CL %S",domain->Scop->Class);
ajFmtPrintSplit(outf,domain->Scop->Fold,"\n# XX\n# FO ",
75," \t\n\r");
ajFmtPrintSplit(outf,domain->Scop->Superfamily,"# XX\n# SF ",
75," \t\n\r");
ajFmtPrintSplit(outf,domain->Scop->Family,"# XX\n# FA ",
75," \t\n\r");
ajFmtPrintF(outf,"# XX\n");
}
else
{
ajFmtPrintF(outf,"# TY CATH\n# XX\n");
ajFmtPrintF(outf,"# CL %S",domain->Cath->Class);
ajFmtPrintSplit(outf,domain->Cath->Architecture,"\n# XX\n# AR ",
75," \t\n\r");
ajFmtPrintSplit(outf,domain->Cath->Topology,"# XX\n# TP ",75,
" \t\n\r");
ajFmtPrintSplit(outf,domain->Cath->Superfamily,"# XX\n# SF ",75,
" \t\n\r");
ajFmtPrintF(outf,"# XX\n");
}
if((domain->Type == ajSCOP))
{
if(noden==1)
ajFmtPrintF(outf,"# SI %d\n# XX",domain->Scop->Sunid_Class);
else if(noden==2)
ajFmtPrintF(outf,"# SI %d\n# XX",domain->Scop->Sunid_Fold);
else if(noden==3)
ajFmtPrintF(outf,"# SI %d\n# XX",
domain->Scop->Sunid_Superfamily);
else if(noden==4)
ajFmtPrintF(outf,"# SI %d\n# XX",
domain->Scop->Sunid_Family);
else
ajFatal("Node number error in domainalign_ProcessStampFile");
}
else
{
if(noden==5)
ajFmtPrintF(outf,"# SI %d\n# XX", domain->Cath->Class_Id);
else if(noden==6)
ajFmtPrintF(outf,"# SI %d\n# XX", domain->Cath->Arch_Id);
else if(noden==7)
ajFmtPrintF(outf,"# SI %d\n# XX",domain->Cath->Topology_Id);
else if(noden==8)
ajFmtPrintF(outf,"# SI %d\n# XX",
domain->Cath->Superfamily_Id);
else if(noden==9)
ajFmtPrintF(outf,"# SI %d\n# XX",domain->Cath->Family_Id);
else
ajFatal("Node number error in domainalign_ProcessStampFile");
}
while((ajReadlineTrim(inf,&line)))
{
/* Increment counter for block of file. */
if((ajStrGetCharPos(line, 1)=='\0'))
{
blk++;
if(blk==4)
blk=1;
continue;
}
/* Block of numbering line and protein sequences. */
if(blk==1)
{
/* Print the number line out as it is. */
if(ajStrPrefixC(line,"Number"))
ajFmtPrintF(outf,"\n# %7s %S\n"," ", line);
else
{
/* Read only the 7 characters
of the domain identifier
code in. */
ajFmtScanS(line, "%S", &temp1);
ajStrAssignSubS(&temp2, temp1, 0, 6);
/* Read the sequence. */
ajStrAssignSubS(&temp3, line, 13, 69);
ajStrExchangeSetCC(&temp3, " ", "X");
ajFmtPrintF(logf, "Replaced ' ' in STAMP alignment "
"with 'X'\n");
ajStrFmtUpper(&temp3);
/* Write domain id code and sequence out. */
ajFmtPrintF(outf,"%-15S%7d %S%7d\n",
temp2, 0, temp3, 0);
}
}
/* Secondary structure filled with '????' (unwanted). */
else if(blk==2)
{
continue;
}
/* Similarity lines. */
else
{
if(ajStrPrefixC(line,"Post"))
{
/* Read the sequence. */
ajStrAssignSubS(&temp3, line, 13, 69);
/* Write post similar line out. */
ajFmtPrintF(outf,"%-15s%7s %S\n","# Post_similar", " ",
temp3);
}
/* Ignore Very and Less similar lines. */
else continue;
}
}
}
else /* ok == ajFalse. */
{
ajWarn("\n***********************************************\n"
"* STAMP was called but output file was EMPTY! *\n"
"* NO OUTPUT FILE GENERATED FOR THIS NODE. *\n"
"***********************************************\n");
ajFmtPrintF(logf, "STAMP called but output file empty. "
"No output file for this node!");
}
/* Clean up and close input and output files. */
ajFileClose(&outf);
ajFileClose(&inf);
ajStrDel(&line);
ajStrDel(&temp1);
ajStrDel(&temp2);
ajStrDel(&temp3);
/* All done. */
return;
}
