int main(int argc, char *argv[])
{
short out2stdout=0;
hashtable ht=new_hashtable(HASHSIZE);
bamFile in,in2;
bamFile out;
int paired;//1 if not paired or pair read 1, 2 otherwise
index_mem=sizeof(hashtable)*sizeof(hashnode**)*HASHSIZE*2;
if (argc != 3) {
fprintf(stderr, "Usage: bam_fix_NH <in.bam> <out.bam or - for stdout>\n");
return 1;
}
// Open file and exit if error
in = bam_open(argv[1], "rb");
out2stdout = strcmp(argv[2], "-")? 0 : 1;
out = strcmp(argv[2], "-")? bam_open(argv[2], "w") : bam_dopen(fileno(stdout), "w");
if (in == 0 ) {
fprintf(stderr, "ERROR: Fail to open BAM file %s\n", argv[1]);
return 1;
}
if (out == 0) {
fprintf(stderr, "ERROR: Fail to open BAM file %s\n", argv[2]);
return 1;
}
unsigned long num_alns=0;
int ref;
// ***********
// Copy header
bam_header_t *header;
header = bam_header_read(in);
bam_header_write(out,header);
// sorted by name?
// Should not rely on the value in SO
bam1_t *aln=bam_init1();
bam1_t *prev=bam_init1();
if (!out2stdout) {
fprintf(stderr,"bam_fix_NH version %s\n",VERSION);
fprintf(stderr,"Processing %s\n",argv[1]);
fprintf(stderr,"Hashing...\n");fflush(stderr);
}
while(bam_read1(in,aln)>=0) { // read alignment
if (aln->core.tid < 0) continue;//ignore unaligned reads
if (aln->core.flag & BAM_FUNMAP) continue;
if (aln->core.flag & BAM_FREAD2) paired=2;
else paired=1;
++num_alns;
new_read_aln(ht,fix_read_name(bam1_qname(aln),paired));
if(!out2stdout) PRINT_ALNS_PROCESSED(num_alns);
}
bam_close(in);
if(!out2stdout) {
fprintf(stderr,"%s%lu\n",BACKLINE,num_alns);
fprintf(stderr,"Hashing complete (%lu alignments)\n",num_alns);
fprintf(stderr,"Memory used: %ld MB\n",index_mem/1024/1024);
fprintf(stderr,"Updating entries with NH and printing BAM...\n");
fflush(stderr);
}
// reopen
in2 = bam_open(argv[1], "rb");
if (in2 == 0 ) {
fprintf(stderr, "ERROR: Fail to open BAM file %s\n", argv[1]);
return 1;
}
header = bam_header_read(in2);
num_alns=0;
while(bam_read1(in2,aln)>=0) { // read alignment
paired=1;
if (aln->core.tid < 0) continue;//ignore unaligned reads
if (aln->core.flag & BAM_FUNMAP) continue;
if (aln->core.flag & BAM_FREAD2) paired=2;
++num_alns;
READ_ALN *r=get_read_aln(ht,fix_read_name(bam1_qname(aln),paired));
assert(r!=NULL);
// update the NH field
uint8_t *old_nh = bam_aux_get(aln, "NH");
int32_t nh=r->ctr;
if (old_nh) {
if (nh!=bam_aux2i(old_nh)) {
fprintf(stderr,"warning: value mismatch! replacing>%s %d->%d\n",bam1_qname(aln),bam_aux2i(old_nh),nh);
}
bam_aux_del(aln, old_nh);
bam_aux_append(aln, "NH", 'i', 4, (uint8_t*)&nh);
#ifdef DEBUG
// printf("!>%s %d\n",bam1_qname(aln),r->ctr);
#endif
}
if (!old_nh) { // add NH
bam_aux_append(aln, "NH", 'i', 4, (uint8_t*)&nh);
#ifdef DEBUG
fprintf(stderr,"!>%s %d\n",bam1_qname(aln),bam_aux2i(old_nh));
#endif
}
bam_write1(out,aln);
if(!out2stdout) PRINT_ALNS_PROCESSED(num_alns);
}
//
bam_destroy1(aln);
bam_close(in2);
bam_close(out);
if(!out2stdout) {
fprintf(stderr,"%s%lu\n",BACKLINE,num_alns);
fprintf(stderr,"Done.\n");
}
return 0;
}
int main(int argc, char *argv[])
{
bamFile in;
sqlite3 * db;
sqlite3_stmt * stmt;
char * sErrMsg = NULL;
char * tail = 0;
int nRetCode;
char sSQL [BUFFER_SIZE] = "\0";
char database[BUFFER_SIZE];
clock_t startClock,startClock2;
if (argc != 2) {
fprintf(stderr, "Usage: bamRindex <in.bam>\n");
return 1;
}
// Open file and exit if error
//in = strcmp(argv[1], "-")? bam_open(argv[1], "rb") : bam_dopen(fileno(stdin), "rb");
//fprintf(stderr,"Options ok\n");
in = bam_open(argv[1], "rb");
if (in == 0 ) {
fprintf(stderr, "ERROR: Fail to open BAM file %s\n", argv[1]);
return 1;
}
//fprintf(stderr,"BAM opened\n");
assert(strcpy(database,argv[1])!=NULL);
assert(strcat(database,".ridx")!=NULL);
remove(database);
// ***********
// Read header
bam_header_t *header;
header = bam_header_read(in);
// sorted by name?
// Should not rely on the value in SO
bam1_t *aln=bam_init1();
unsigned long num_alns=0;
/*********************************************/
/* Open the Database and create the Schema */
// TODO: check the errors
sqlite3_open(database, &db);
sqlite3_exec(db, TABLE, NULL, NULL, &sErrMsg); // create the table
SQLITE_CHECK_ERROR();
startClock = clock();
sqlite3_exec(db, "PRAGMA synchronous = 0;", NULL, NULL, &sErrMsg);
SQLITE_CHECK_ERROR();
sqlite3_exec(db, "PRAGMA journal_mode = OFF;", NULL, NULL, &sErrMsg);
SQLITE_CHECK_ERROR();
// Use up to 8GB of memory
sqlite3_exec(db, "PRAGMA cache_size = -8000000;", NULL, NULL, &sErrMsg);
SQLITE_CHECK_ERROR();
sqlite3_exec(db, "BEGIN TRANSACTION;", NULL, NULL, &sErrMsg);
SQLITE_CHECK_ERROR();
while(bam_read1(in,aln)>=0) { // read alignment
//aln->core.tid < 0 ?
uint8_t *nh = bam_aux_get(aln, "NH");
uint8_t *nm = bam_aux_get(aln, "NM");
uint8_t *xs = bam_aux_get(aln, "XS");
BOOLEAN isPrimary;
BOOLEAN isMapped;
BOOLEAN notMapped;
BOOLEAN isDuplicate;
BOOLEAN isNotPassingQualityControls;
BOOLEAN isPaired;
BOOLEAN isSecondMateRead,isProperPair;
//secondary alignment
notMapped=(aln->core.flag & BAM_FUNMAP) ? TRUE: FALSE;
//notMapped=((aln->core.flag & BAM_FUNMAP) || (aln->core.mtid ==0)) ? TRUE: FALSE;
isMapped=!notMapped;
isPrimary= (aln->core.flag & BAM_FSECONDARY) ? FALSE:TRUE;
isProperPair=(aln->core.flag & BAM_FPROPER_PAIR) ? TRUE:FALSE;
isPaired=(aln->core.flag & BAM_FPAIRED ) ? TRUE:FALSE;
isSecondMateRead=(aln->core.flag & BAM_FREAD2 ) ? TRUE: FALSE;
isNotPassingQualityControls=(aln->core.flag & BAM_FQCFAIL ) ? TRUE:FALSE;
isDuplicate=(aln->core.flag & BAM_FDUP) ? TRUE: FALSE;
BOOLEAN isSpliced=FALSE;
BOOLEAN hasSimpleCigar=TRUE;
int nSpliced=0;
int i;
if (aln->core.n_cigar != 0) {
for (i = 0; i < aln->core.n_cigar; ++i) {
char l="MIDNSHP=X"[bam1_cigar(aln)[i]&BAM_CIGAR_MASK];
//fprintf(stderr,"%c",l);
if ( l == 'N' ) { isSpliced=TRUE; hasSimpleCigar=FALSE;++nSpliced;}
if ( l != 'M' && l!='=' ) { hasSimpleCigar=FALSE;}
}
}
//fprintf(stderr,"read %ld\n",num_alns);
// isDuplicate,isNotPassingQualityControls,
// isSpliced,isPAired,isPrimary,hasSimpleCigar,isSecondMateRead,isProperPair,nh,nm,qual/mapq,xs
sprintf(sSQL,"INSERT into bam_index values (%d,%d,%d,%d,%d,%d,%d,%d,%d,%d,%d,%d,'%c')",
isDuplicate,isNotPassingQualityControls,
nSpliced,isPaired,isPrimary,isMapped,hasSimpleCigar,isSecondMateRead,isProperPair,
(nh==0?0:bam_aux2i(nh)),(nm==0?0:bam_aux2i(nm)),
aln->core.qual,
(xs==0?' ':(bam_aux2A(xs)==0?' ':bam_aux2A(xs))));
sqlite3_exec(db, sSQL, NULL, NULL, &sErrMsg);
SQLITE_CHECK_ERROR();
++num_alns;
PRINT_ALNS_PROCESSED(num_alns);
}
bam_close(in);
sqlite3_exec(db, "END TRANSACTION;", NULL, NULL, &sErrMsg);
SQLITE_CHECK_ERROR();
printf("\nImported %d records in %4.2f seconds\n", num_alns, ( (double) (clock() - startClock))/CLOCKS_PER_SEC);
// Create the indexes
startClock2 = clock();
// generating the indexes does not pay off
//sqlite3_exec(db, INDEXES, NULL, NULL, &sErrMsg);
//printf("Indexed %d records in %4.2f seconds\n", num_alns, ( (double) (clock() - startClock2))/CLOCKS_PER_SEC);
printf("Total time: %4.2f seconds\n", ((double)(clock() - startClock))/CLOCKS_PER_SEC);
sqlite3_close(db);
return 0;
}
int main(int argc, char *argv[])
{
short out2stdout=0;
bamFile in,in2;
bamFile out;
if (argc != 3) {
fprintf(stderr, "Usage: bam_fix_se_flag <in.bam> <out.bam or - for stdout>\n");
return 1;
}
// Open file and exit if error
in = bam_open(argv[1], "rb");
out2stdout = strcmp(argv[2], "-")? 0 : 1;
out = strcmp(argv[2], "-")? bam_open(argv[2], "w") : bam_dopen(fileno(stdout), "w");
if (in == 0 ) {
fprintf(stderr, "ERROR: Fail to open BAM file %s\n", argv[1]);
return 1;
}
if (out == 0) {
fprintf(stderr, "ERROR: Fail to open BAM file %s\n", argv[2]);
return 1;
}
unsigned long num_alns=0;
int ref;
// ***********
// Copy header
bam_header_t *header;
header = bam_header_read(in);
bam_header_write(out,header);
// sorted by name?
// Should not rely on the value in SO
bam1_t *aln=bam_init1();
bam1_t *prev=bam_init1();
if (!out2stdout) {
fprintf(stderr,"bam_fix_se_flag version %s\n",VERSION);
fprintf(stderr,"Processing %s\n",argv[1]);
}
// reopen
in2 = bam_open(argv[1], "rb");
if (in2 == 0 ) {
fprintf(stderr, "ERROR: Fail to open BAM file %s\n", argv[1]);
return 1;
}
header = bam_header_read(in2);
num_alns=0;
while(bam_read1(in2,aln)>=0) { // read alignment
if (aln->core.tid < 0) continue;//ignore unaligned reads
if (aln->core.flag & BAM_FUNMAP) continue;
if (aln->core.flag & BAM_FPAIRED ) { // PAIRED
} else { //SE
//turn off the other pair related flags
aln->core.flag&=~BAM_FPROPER_PAIR;
aln->core.flag&=~BAM_FMUNMAP;
aln->core.flag&=~BAM_FREAD1;
aln->core.flag&=~BAM_FREAD2;
fprintf(stderr, ".");
}
bam_write1(out,aln);
if(!out2stdout) PRINT_ALNS_PROCESSED(num_alns);
++num_alns;
}
//
bam_destroy1(aln);
bam_close(in2);
bam_close(out);
if(!out2stdout) {
fprintf(stderr,"%s%lu\n",BACKLINE,num_alns);
fprintf(stderr,"Done.\n");
}
return 0;
}