Sequence * init_SequenceDB(SequenceDB * sdb,int * return_status)
{
sdb->current_source = 0;
load_next_fs_SequenceDB(sdb);
return reload_SequenceDB(NULL,sdb,return_status);
}
ThreeStateModel * read_TSM_ThreeStateDB(ThreeStateDB * mdb,int * return_status)
{
ThreeStateModel * tsm;
Protein * pro;
Sequence * seq;
if( mdb->hmm_model_end != -1 && mdb->current_no == mdb->hmm_model_end ) {
*return_status = DB_RETURN_END;
return NULL;
}
mdb->current_no++;
switch( mdb->dbtype ) {
case TSMDB_SINGLE :
*return_status = DB_RETURN_END;
if( mdb->single->rm == NULL ) {
warn("Threestate model without an internal random model!");
mdb->single->rm = hard_link_RandomModel(mdb->rm);
}
return hard_link_ThreeStateModel(mdb->single);
case TSMDB_HMMER1PFAM :
tsm= read_next_TSM_PfamHmmer1DB(mdb->phdb,return_status);
set_startend_policy_ThreeStateModel(tsm,mdb->type,30,0.2);
return tsm;
case TSMDB_PROTEIN :
if( mdb->seq_cache != NULL ) {
/* just after an open. Should actually use this sequence, and flush the cache */
pro = Protein_from_Sequence(hard_link_Sequence(mdb->seq_cache));
mdb->seq_cache = free_Sequence(mdb->seq_cache);
*return_status = DB_RETURN_OK;
} else {
/* reload a sequence from a database */
seq = reload_SequenceDB(NULL,mdb->sdb,return_status);
/* exit now if error */
if( *return_status == DB_RETURN_ERROR ) {
return NULL; /* might have leaked memory. Ugh! */
}
/* if we get NULL... for the moment, silent flag end */
if( seq == NULL ) {
*return_status = DB_RETURN_END;
return NULL;
}
pro = Protein_from_Sequence(seq);
}
if( pro == NULL ) {
warn("Could not convert sequence to a protein. Exiting!");
*return_status = DB_RETURN_ERROR;
return NULL;
}
/* convert protein to threestatemodel */
tsm = ThreeStateModel_from_half_bit_Sequence(pro,mdb->comp,mdb->rm,mdb->gap,mdb->ext);
if( tsm == NULL ) {
warn("Could not convert protein to threestatemode. Exiting!");
free_Protein(pro);
*return_status = DB_RETURN_ERROR;
return NULL;
}
/* DB status already set by seqdb */
return tsm;
case TSMDB_GENERIC :
tsm = ((*mdb->reload_generic)(mdb,return_status));
if( tsm == NULL ) {
return NULL; /* means end of database */
}
set_startend_policy_ThreeStateModel(tsm,mdb->type,30,0.2);
return tsm;
default :
warn("Got an unrecognisable tsm db type in read-load");
return NULL;
}
}
ComplexSequence * reload_cDNADB(ComplexSequence * last,cDNADB * cdnadb,int * return_status)
{
ComplexSequence * cs;
Sequence * seq,*temp;
/** free Complex Sequence **/
if ( last != NULL ) {
free_ComplexSequence(last);
}
if( cdnadb->forward_only == TRUE) {
temp = reload_SequenceDB(NULL,cdnadb->sdb,return_status);
if ( *return_status != DB_RETURN_OK ) {
return NULL;
}
cs = new_ComplexSequence(temp,cdnadb->cses);
return cs;
}
if( cdnadb->is_single_seq == TRUE ) {
if( cdnadb->done_forward == TRUE ) {
*return_status = DB_RETURN_OK;
cdnadb->done_forward = FALSE;
return hard_link_ComplexSequence(cdnadb->rev);
} else {
*return_status = DB_RETURN_END;
return NULL;
}
}
/** standard database **/
if( cdnadb->done_forward == TRUE ) {
if( cdnadb->current == NULL ) {
warn("A bad internal cDNA db error - unable to find current sequence in db reload");
*return_status = DB_RETURN_ERROR;
return NULL;
}
temp = reverse_complement_Sequence(cdnadb->current);
if( temp == NULL ) {
warn("A bad internal cDNA db error - unable to reverse complements current");
*return_status = DB_RETURN_ERROR;
return NULL;
}
cs = new_ComplexSequence(temp,cdnadb->cses);
if( cs == NULL ) {
warn("A bad internal cDNA db error - unable to make complex sequence in db reload");
*return_status = DB_RETURN_ERROR;
return NULL;
}
free_Sequence(temp);
cdnadb->done_forward = FALSE;
return cs;
}
/* otherwise we have to get a new sequence */
seq = reload_SequenceDB(NULL,cdnadb->sdb,return_status);
if( seq == NULL || *return_status == DB_RETURN_ERROR || *return_status == DB_RETURN_END ) {
return NULL; /** error already reported **/
}
uppercase_Sequence(seq);
if( force_to_dna_Sequence(seq,cdnadb->error_tol,NULL) == FALSE ) {
if( cdnadb->error_handling == CDNADB_READ_THROUGH ) {
warn("Unable to map %s sequence to a cDNA sequence, but ignoring that for the moment...",seq->name);
free_Sequence(seq);
return reload_cDNADB(NULL,cdnadb,return_status);
} else {
warn("Unable to map %s sequence to a cDNA sequence. Failing",seq->name);
*return_status = DB_RETURN_ERROR;
return NULL;
}
}
cs = new_ComplexSequence(seq,cdnadb->cses);
if( cs == NULL ) {
if( cdnadb->error_handling == CDNADB_READ_THROUGH ) {
warn("Unable to map %s sequence to a cDNA sequence, but ignoring that for the moment...",seq->name);
free_Sequence(seq);
return reload_cDNADB(NULL,cdnadb,return_status);
} else {
warn("Unable to map %s sequence to a cDNA sequence. Failing",seq->name);
*return_status = DB_RETURN_ERROR;
return NULL;
}
}
cdnadb->current = free_Sequence(cdnadb->current);
cdnadb->current = seq;
cdnadb->done_forward= TRUE;
return cs;
}
ComplexSequence * reload_GenomicDB(ComplexSequence * last,GenomicDB * gendb,int * return_status)
{
ComplexSequence * cs;
Sequence * seq;
Genomic *temp;
Genomic * gen;
/** NB - notice that we don't do silly things with free's. Maybe we should **/
if( gendb->is_single_seq == TRUE ) {
if( gendb->done_forward == TRUE ) {
*return_status = DB_RETURN_OK;
gendb->done_forward = FALSE;
return hard_link_ComplexSequence(gendb->rev);
} else {
*return_status = DB_RETURN_END;
return NULL;
}
}
/** standard database **/
/** free Complex Sequence **/
if ( last != NULL ) {
free_ComplexSequence(last);
}
if( gendb->done_forward == TRUE ) {
if( gendb->current == NULL ) {
warn("A bad internal genomic db error - unable to find current sequence in db reload");
*return_status = DB_RETURN_ERROR;
return NULL;
}
temp = reverse_complement_Genomic(gendb->current);
if( temp == NULL ) {
warn("A bad internal genomic db error - unable to reverse complements current");
*return_status = DB_RETURN_ERROR;
return NULL;
}
cs = evaluate_ComplexSequence_Genomic(temp,gendb->cses,0,gendb->repeat_in_cds_score);
if( cs == NULL ) {
warn("A bad internal genomic db error - unable to make complex sequence in db reload");
*return_status = DB_RETURN_ERROR;
return NULL;
}
free_Genomic(temp);
gendb->done_forward = FALSE;
return cs;
}
/* otherwise we have to get a new sequence */
seq = reload_SequenceDB(NULL,gendb->sdb,return_status);
if( seq == NULL || *return_status == DB_RETURN_ERROR || *return_status == DB_RETURN_END ) {
return NULL; /** error already reported **/
}
uppercase_Sequence(seq);
/* check dna status. We assumme someone knows what he is doing when he makes a genomic db!*/
if( seq->type != SEQUENCE_DNA) {
warn("Sequence from %s data entry doesn't look like DNA. Forcing it to",seq->name);
}
force_to_dna_Sequence(seq,1.0,NULL);
if( force_to_dna_Sequence(seq,0.1,NULL) == FALSE ) {
if( gendb->error_handling == GENDB_READ_THROUGH ) {
warn("Unable to map %s sequence to a genomic sequence, but ignoring that for the moment...",seq->name);
free_Sequence(seq);
return reload_GenomicDB(NULL,gendb,return_status);
} else {
warn("Unable to map %s sequence to a genomic sequence. Failing",seq->name);
*return_status = DB_RETURN_ERROR;
return NULL;
}
}
gen = Genomic_from_Sequence_Nheuristic(seq,gendb->length_of_N);
cs = evaluate_ComplexSequence_Genomic(gen,gendb->cses,0,gendb->repeat_in_cds_score);
if( cs == NULL ) {
if( gendb->error_handling == GENDB_READ_THROUGH ) {
warn("Unable to map %s sequence to a genomic sequence, but ignoring that for the moment...",seq->name);
free_Sequence(seq);
return reload_GenomicDB(NULL,gendb,return_status);
} else {
warn("Unable to map %s sequence to a genomic sequence. Failing",seq->name);
*return_status = DB_RETURN_ERROR;
return NULL;
}
}
gendb->current = free_Genomic(gendb->current);
gendb->current = gen;
gendb->done_forward= TRUE;
return cs;
}
