int read_dssp (Options * options, Protein *protein) {
char line[LONGSTRING] = {'\0'};
char pdb_id[PDB_ATOM_RES_NO_LEN+2] = {'\0'};
char tmp[5];
int resctr, acc, total = 0;
FILE *fptr;
fptr = efopen(options->dssp_file_name,"r");
while( fgets( line, LONGSTRING, fptr) != NULL){
if( ! strncmp(line+5,"RESIDUE", 7)){
break;
}
}
while( fgets( line, LONGSTRING, fptr) != NULL){
if ( strchr ( line, '!') ) continue;
strncpy ( pdb_id, line+6, 5);
string_clean (pdb_id, 5);
if ( ! pdb_id[0] ) continue;
strncpy ( tmp, line+34, 4);
tmp[4] = '\0';
acc = atoi(tmp);
if ( acc > options->acc_cutoff ) {
for ( resctr=0; resctr < protein->length; resctr++ ) {
if ( !strcmp (pdb_id, protein->sequence[resctr].pdb_id) ) {
protein->sequence[resctr].solvent_accessible = 1;
total++;
}
}
}
}
fclose (fptr);
if ( ! total ) {
fprintf (stderr, "No residue solvent accessible (?!)\n");
return 1;
}
return 0;
}
int read_pdb ( char * pdbname, Protein * protein, char chain) {
Residue * sequence;
FILE * fptr = NULL;
char line[BUFFLEN];
char oldresno[PDB_ATOM_RES_NO_LEN+2]; /* res name: 4 digits + insertion code + \0 */
char oldrestype [PDB_ATOM_RES_NAME_LEN+2];
char tmp[PDB_ATOM_X_LEN+1], *auxptr;
int atomctr, resctr, no_res, ctr, nonblank;
char single_letter ( char code[]);
/* open file */
fptr = fopen ( pdbname, "r");
if ( !fptr ) {
fprintf (stderr, "Cno %s.\n", pdbname);
return 1;
}
/* warn if no chain given */
if ( !chain) fprintf ( stderr,"No chain specified. Using the first one.\n");
/* count residues */
memset (line, 0, BUFFLEN);
memset (oldresno, 0, PDB_ATOM_RES_NO_LEN+2);
resctr = 0;
while(fgets(line, BUFFLEN, fptr)!=NULL){
if ( chain && line[PDB_ATOM_CHAINID] != chain ) continue;
if ( ! strncmp(line,"TER", 3) || ! strncmp(line,"END", 3) ) break;
if( ! strncmp(line,"ATOM", 4) || ! strncmp(line,"HETATM", 6)){
if ( strncmp (line+PDB_ATOM_RES_NO, oldresno, PDB_ATOM_RES_NO_LEN+1) ) {
strncpy (oldresno, line+PDB_ATOM_RES_NO, PDB_ATOM_RES_NO_LEN+1);
oldresno[PDB_ATOM_RES_NO_LEN+1] = '\0';
/* printf ( "New residue number: %s \n", oldresno); */
resctr ++;
}
}
}
no_res = resctr;
/* printf ("no residues: %d\n", no_res); */
/* allocate space */
sequence = NULL;
sequence = emalloc ( no_res*sizeof (Residue));
/* read in the atom */
rewind ( fptr);
memset (line, 0, BUFFLEN);
memset (oldresno, 0, PDB_ATOM_RES_NO_LEN+2);
/* tyring to account for the insertion code */
memset (oldrestype, 0, PDB_ATOM_RES_NAME_LEN+2);
resctr= -1;
atomctr = 0;
while(fgets(line, BUFFLEN, fptr)!=NULL){
if ( chain && line[PDB_ATOM_CHAINID] != chain ) continue;
if ( ! strncmp(line,"TER", 3) || ! strncmp(line,"END", 3) ) break;
if( ! strncmp(line,"ATOM", 4) || ! strncmp(line,"HETATM", 6)){
/* if it's a hydrogen - skip */
if ( line[PDB_ATOM_ATOM_NAME] == 'H'
|| line[PDB_ATOM_ATOM_NAME+1] == 'H') continue;
/* adjust the counters */
if ( strncmp (line+PDB_ATOM_RES_NO, oldresno, PDB_ATOM_RES_NO_LEN+1) ) {
strncpy (oldresno, line+PDB_ATOM_RES_NO, PDB_ATOM_RES_NO_LEN+1);
strncpy (oldrestype, line+PDB_ATOM_RES_NAME, PDB_ATOM_RES_NAME_LEN);
oldresno[PDB_ATOM_RES_NO_LEN+1] = '\0';
oldrestype[PDB_ATOM_RES_NAME_LEN] = '\0';
resctr ++;
atomctr = 0;
sequence[resctr].no_atoms = 1;
strncpy ( sequence[resctr].pdb_id, oldresno, PDB_ATOM_RES_NO_LEN+2);
sequence[resctr].pdb_id[PDB_ATOM_RES_NO_LEN+1] = '\0';
strncpy ( sequence[resctr].res_type, oldrestype, PDB_ATOM_RES_NAME_LEN+1);
sequence[resctr].res_type[PDB_ATOM_RES_NAME_LEN] = '\0';
sequence[resctr].res_type_short = single_letter ( sequence[resctr].res_type );
if ( !sequence[resctr].res_type_short ) return 1;
} else {
atomctr ++;
sequence[resctr].no_atoms = atomctr + 1;
if ( atomctr >= MAX_NO_ATOMS ) {
fprintf ( stderr, "Error: I thought every aa has < %d atoms.\n",
MAX_NO_ATOMS );
return 1;
}
}
/* read in atom info */
auxptr = line+ PDB_ATOM_ATOM_NAME;
memset ( tmp, 0, PDB_ATOM_ATOM_NAME_LEN+1);
/* skip initial blanks*/
ctr = 0;
while ( !(isalpha (*(auxptr + ctr))) && (ctr <= PDB_ATOM_ATOM_NAME_LEN) ) ctr++;
/* copy alphanum info */
nonblank = 0;
while ( isalpha (*(auxptr +ctr)) && (ctr <= PDB_ATOM_ATOM_NAME_LEN) ) {
tmp[nonblank] = *(auxptr +ctr);
nonblank ++;
ctr++;
}
strncpy ( sequence[resctr].atom[atomctr].type, tmp, PDB_ATOM_ATOM_NAME_LEN );
/* is this a backbone atom?*/
sequence[resctr].atom[atomctr].backbone = 0;
if ( nonblank == 1) {
sequence[resctr].atom[atomctr].backbone =
!( strcmp ( tmp, "N") && strcmp ( tmp, "C") && strcmp ( tmp, "O") );
} else if ( nonblank == 2) {
sequence[resctr].atom[atomctr].backbone = ! strcmp ( tmp, "CA" );
}
/* printf ( " %4d %4d %4s is backbone: %1d \n", resctr, atomctr, */
/* sequence[resctr].atom[atomctr].type, sequence[resctr].atom[atomctr].backbone); */
strncpy ( tmp, line+PDB_ATOM_X, PDB_ATOM_X_LEN);
tmp[PDB_ATOM_X_LEN] = '\0';
sequence[resctr].atom[atomctr].x=atof(tmp);
strncpy ( tmp, line+PDB_ATOM_Y, PDB_ATOM_Y_LEN);
tmp[PDB_ATOM_Y_LEN] = '\0';
sequence[resctr].atom[atomctr].y=atof(tmp);
strncpy ( tmp, line+PDB_ATOM_Z, PDB_ATOM_Z_LEN);
tmp[PDB_ATOM_Z_LEN] = '\0';
sequence[resctr].atom[atomctr].z=atof(tmp);
}
}
/* close file */
fclose (fptr);
/* clean PDB id tags from spaces */
for (resctr=0; resctr < no_res; resctr ++ ) {
string_clean (sequence[resctr].pdb_id, PDB_ATOM_RES_NO_LEN+2);
}
/*return values: */
protein->sequence= sequence;
protein->length = no_res;
return 0;
}
/*
MAIN()
------
*/
int main(int argc, char *argv[])
{
static char *seperators = " ";
char *file, *token, *where_to, *filename; // *start;
char **term_list, **first, **last, **current;
ANT_link_extract_term *link_index, *index_term;
long terms_in_index, current_docid, param, file_number;
long lowercase_only, first_param;
long is_utf8_token, cmp, is_substring = FALSE; // token_len
char *command;
ANT_directory_iterator_object file_object;
char buffer[1024 * 1024];
if (argc < 3)
exit(printf("Usage:%s [-chinese] [-lowercase] <index> <file_to_link> ...\n", argv[0]));
first_param = 1;
lowercase_only = FALSE;
chinese = FALSE;
for (param = 1; param < argc; param++)
{
if (*argv[param] == '-')
{
command = argv[param] + 1;
if (strcmp(command, "lowercase") == 0)
{
lowercase_only = TRUE;
++first_param;
}
else if (strcmp(command, "chinese") == 0)
{
chinese = TRUE;
++first_param;
}
else
exit(printf("Unknown parameter:%s\n", argv[param]));
}
}
link_index = read_index(argv[first_param], &terms_in_index);
file_number = 1;
for (param = first_param + 1; param < argc; param++)
{
ANT_directory_iterator_recursive disk(argv[param]); // make the recursive pattern matching as for default files reading
if (disk.first(&file_object) == NULL)
file = filename = NULL;
else
{
filename = file_object.filename;
file = ANT_disk::read_entire_file(filename);
}
while (file != NULL)
{
current_docid = get_doc_id(file);
if (current_docid > 0)
{
// printf("ID:%d\n", current_docid);
string_clean(file, lowercase_only, TRUE);
current = term_list = new char *[strlen(file)]; // this is the worst case by far
if (chinese)
create_utf8_token_list(file, term_list);
else
{
for (token = strtok(file, seperators); token != NULL; token = strtok(NULL, seperators))
*current++ = token;
*current = NULL;
}
for (first = term_list; *first != NULL; first++)
{
// fprintf(stderr, "%s\n", *first);
where_to = buffer;
for (last = first; *last != NULL; last++)
{
if (where_to == buffer)
{
strcpy(buffer, *first);
where_to = buffer + strlen(buffer);
if (chinese)
{
if ((*first[0] & 0x80) && isutf8(*first))
is_utf8_token = TRUE;
else
is_utf8_token = FALSE;
}
}
else
{
if (!chinese)
*where_to++ = ' ';
strcpy(where_to, *last);
where_to += strlen(*last);
}
*where_to = '\0';
index_term = find_term_in_list(buffer, link_index, terms_in_index);
if (index_term == NULL)
break; // we're after the last term in the list so can stop because we can't be a substring
if (chinese)
{
is_substring = FALSE;
cmp = utf8_token_compare(buffer, index_term->term, &is_substring);
}
else
cmp = string_compare(buffer, index_term->term);
if (cmp == 0) // we're a term in the list
{
index_term->total_occurences++;
if (index_term->last_docid != current_docid)
{
index_term->last_docid = current_docid;
index_term->docs_containing_term++;
}
}
else
{
if (chinese)
cmp = is_substring == TRUE ? 0 : 1;
else
cmp = memcmp(buffer, index_term->term, strlen(buffer));
if (cmp != 0)
break; // we're a not a substring so we can't find a longer term
}
}
}
if (chinese)
free_utf8_token_list(term_list);
delete [] term_list;
delete [] file;
if (file_number % 1000 == 0)
fprintf(stderr, "Files processed:%d\n", file_number);
file_number++;
}
else
fprintf(stderr, "Error reading file %s\n", filename);
//filename = disk.get_next_filename();
if (disk.next(&file_object) == NULL)
file = filename = NULL;
else
{
filename = file_object.filename;
file = ANT_disk::read_entire_file(filename);
}
}
}
print_answer(link_index, terms_in_index);
fprintf(stderr, "%s Completed\n", argv[0]);
return 0;
}
int fill_protein_info ( FILE * fptr, char chain, Protein * protein) {
/* TODO for the moment we rely on PDB annotation
to extract structural elements - that should be changed */
Residue * sequence = NULL;
char line[BUFFLEN];
char oldresno[PDB_ATOM_RES_NO_LEN+2];
/* res name: 4 digits + insertion code + \0 */
char oldrestype [PDB_ATOM_RES_NAME_LEN+2];
char tmp[BUFFLEN], *auxptr;
char atomtypes_read_in[BUFFLEN];
char old_chain;
int atomctr, resctr, no_res,ctr, nonblank;
int retval;
int chain_found;
int ca_trace;
char single_letter ( char code[]);
int has_backbone (Residue * sequence, int from, int to);
/********************************************/
/********************************************/
/* cleanup */
memset (protein, 0, sizeof(Protein) );
/********************************************/
/********************************************/
/* count residues */
memset (line, 0, BUFFLEN);
memset (oldresno, 0, PDB_ATOM_RES_NO_LEN+2);
resctr = 0;
chain_found = 0;
old_chain = '\0';
while(fgets(line, BUFFLEN, fptr)!=NULL){
if (resctr) {
if ( ! strncmp(line,"END", 3) || (chain && line[PDB_ATOM_CHAINID] != old_chain) )
break;
}
if (chain && line[PDB_ATOM_CHAINID] != chain) continue;
chain_found = 1;
if( ! strncmp(line,"ATOM", 4)){
if ( strncmp (line+PDB_ATOM_RES_NO, oldresno, PDB_ATOM_RES_NO_LEN+1) ) {
strncpy (oldresno, line+PDB_ATOM_RES_NO, PDB_ATOM_RES_NO_LEN+1);
oldresno[PDB_ATOM_RES_NO_LEN+1] = '\0';
/* handling the case when the chain is not given, meaning: "take the first chain" */
old_chain = line[PDB_ATOM_CHAINID];
resctr ++;
}
}
}
/* sanity: */
if ( chain && ! chain_found) {
fprintf (stderr, "Chain %c not found.\n", chain);
return ERR_NO_FILE_OR_CHAIN;
}
no_res = resctr;
if ( !no_res ) return -1; /* take it as the end of the read */
/* allocate space */
sequence = emalloc ( no_res*sizeof (Residue));
if ( ! sequence ) return 1;
/*********************************************/
/*********************************************/
/* read in residue numbers and atom coords */
rewind ( fptr);
memset (line, 0, BUFFLEN);
old_chain = '\0';
memset (oldresno, 0, PDB_ATOM_RES_NO_LEN+2);
memset (oldrestype, 0, PDB_ATOM_RES_NAME_LEN+2);
resctr= -1;
atomctr = 0;
ca_trace = 1;
while(fgets(line, BUFFLEN, fptr)!=NULL){
if ( resctr > -1) {
if (! strncmp(line,"END", 3) || (chain && line[PDB_ATOM_CHAINID] != old_chain))
break;
}
if ( chain && line[PDB_ATOM_CHAINID] != chain ) continue;
if( ! strncmp(line,"ATOM", 4) ){
/* if it's a hydrogen - skip */
if ( line[PDB_ATOM_ATOM_NAME] == 'H'
|| line[PDB_ATOM_ATOM_NAME+1] == 'H') continue;
/* adjust the counters */
if ( strncmp (line+PDB_ATOM_RES_NO, oldresno, PDB_ATOM_RES_NO_LEN+1) ) {
/*+1 in PDB_ATOM_RES_NO_LEN+1 means I am including the insertion code
in the identifier */
strncpy (oldresno, line+PDB_ATOM_RES_NO, PDB_ATOM_RES_NO_LEN+1);
strncpy (oldrestype, line+PDB_ATOM_RES_NAME, PDB_ATOM_RES_NAME_LEN);
oldresno[PDB_ATOM_RES_NO_LEN+1] = '\0';
oldrestype[PDB_ATOM_RES_NAME_LEN] = '\0';
/* handling the case when the chain is not given, meaning: "take the first chain" */
old_chain = line[PDB_ATOM_CHAINID];
resctr ++;
if ( resctr >= no_res ) {
fprintf (stderr, "Error reading pdb: resctr:%d no res: %d\n",
resctr, no_res);
return ERR_NONSENSE;
}
atomctr = 0;
/* keep track of atom types we have read in */
memset (atomtypes_read_in, 0, BUFFLEN*sizeof(char));
atomtypes_read_in[0] = '_';
sequence[resctr].no_atoms = 1;
strncpy ( sequence[resctr].pdb_id, oldresno, PDB_ATOM_RES_NO_LEN+2);
sequence[resctr].pdb_id[PDB_ATOM_RES_NO_LEN+1] = '\0';
strncpy ( sequence[resctr].res_type, oldrestype, PDB_ATOM_RES_NAME_LEN+1);
sequence[resctr].res_type[PDB_ATOM_RES_NAME_LEN] = '\0';
sequence[resctr].res_type_short =
single_letter ( sequence[resctr].res_type );
/* modified residues are ok for the purposes here */
/* unless they are sugars or some such - deal with it below */
/* by checking the backbone atoms */
///if ( !sequence[resctr].res_type_short ) return 1;
} else {
atomctr ++;
sequence[resctr].no_atoms = atomctr + 1;
if ( atomctr >= MAX_NO_ATOMS ) {
fprintf ( stderr,
"Error parsing pdb: I thought every aa has < %d atoms.\n",
MAX_NO_ATOMS );
return ERR_MAX_ATOMS;
}
}
/* read in atom info */
auxptr = line+ PDB_ATOM_ATOM_NAME;
memset ( tmp, 0, PDB_ATOM_ATOM_NAME_LEN+1);
/* skip initial blanks*/
ctr = 0;
while ( !(isalpha (*(auxptr + ctr))) &&
(ctr <= PDB_ATOM_ATOM_NAME_LEN) ) ctr++;
/* copy alphanum info */
nonblank = 0;
while ( isalnum (*(auxptr +ctr)) && (ctr <= PDB_ATOM_ATOM_NAME_LEN) ) {
tmp[nonblank] = *(auxptr +ctr);
nonblank ++;
ctr++;
}
/* have we already seen this atom type by any chance? */
tmp[nonblank] = '_';
if ( strstr (atomtypes_read_in, tmp) ) {
/* ahould I check for an alt location code, or just move on? */
//printf ( " %s >> %s //// %s\n", sequence[resctr].pdb_id, atomtypes_read_in, tmp);
continue;
} else {
sprintf (atomtypes_read_in, "%s%s", atomtypes_read_in, tmp);
}
tmp[nonblank] = '\0';
strncpy ( sequence[resctr].atom[atomctr].type, tmp, PDB_ATOM_ATOM_NAME_LEN );
/* is this a backbone atom?*/
sequence[resctr].atom[atomctr].backbone = 0;
if ( nonblank == 1) {
sequence[resctr].atom[atomctr].backbone =
!( strcmp ( tmp, "N") && strcmp ( tmp, "C") && strcmp ( tmp, "O") );
} else if ( nonblank == 2) {
if ( ! strcmp ( tmp, "CA" )) {
sequence[resctr].atom[atomctr].backbone = 1;
sequence[resctr].Ca = sequence[resctr].atom+atomctr;
} else {
sequence[resctr].atom[atomctr].backbone = 0;
}
}
/* check if this is Ca trace */
if ( strcmp ( tmp, "CA" ) ) ca_trace = 0;
strncpy ( tmp, line+PDB_ATOM_X, PDB_ATOM_X_LEN);
tmp[PDB_ATOM_X_LEN] = '\0';
sequence[resctr].atom[atomctr].x=atof(tmp);
strncpy ( tmp, line+PDB_ATOM_Y, PDB_ATOM_Y_LEN);
tmp[PDB_ATOM_Y_LEN] = '\0';
sequence[resctr].atom[atomctr].y=atof(tmp);
strncpy ( tmp, line+PDB_ATOM_Z, PDB_ATOM_Z_LEN);
tmp[PDB_ATOM_Z_LEN] = '\0';
sequence[resctr].atom[atomctr].z=atof(tmp);
}
}
if ( ca_trace) return ERR_SSE_NONE|ERR_CA_TRACE;
/* clean PDB id tags from spaces */
for (resctr=0; resctr < no_res; resctr ++ ) {
retval = string_clean (sequence[resctr].pdb_id, PDB_ATOM_RES_NO_LEN+1);
if ( retval ) {
fprintf (stderr, "Error in read_pdb(): empty id string for residue with sequential no %d.\n", resctr);
return ERR_NONSENSE;
}
}
/* store the sequence and its length */
protein->sequence = sequence;
protein->length = no_res;
return 0;
}