static int isHomologous (Array kgTreeFams, char *transcript1, char *transcript2)
{
Texta tokens;
int i,j;
char *treeFamId;
static Texta treeFamIdsTranscript1 = NULL;
static Texta treeFamIdsTranscript2 = NULL;
textCreateClear (treeFamIdsTranscript1,100);
textCreateClear (treeFamIdsTranscript2,100);
tokens = textFieldtokP (transcript1,"|");
for (i = 0; i < arrayMax (tokens); i++) {
if (treeFamId = lookUpTreeFam (kgTreeFams,textItem (tokens,i))) {
textAdd (treeFamIdsTranscript1,treeFamId);
}
}
textDestroy (tokens);
tokens = textFieldtokP (transcript2,"|");
for (i = 0; i < arrayMax (tokens); i++) {
if (treeFamId = lookUpTreeFam (kgTreeFams,textItem (tokens,i))) {
textAdd (treeFamIdsTranscript2,treeFamId);
}
}
textDestroy (tokens);
for (i = 0; i < arrayMax (treeFamIdsTranscript1); i++) {
for (j = 0; j < arrayMax (treeFamIdsTranscript2); j++) {
if (strEqual (textItem (treeFamIdsTranscript1,i),textItem (treeFamIdsTranscript2,j))) {
return 1;
}
}
}
return 0;
}
int checkOriginal ( BlatQuery* blQ, PslEntry* entry )
{
int res = 0;
Texta t = textFieldtokP( blQ->qName, "|");
if( ( strEqual ( textItem( t, 1 ), entry->tName ) ) &&
( atoi( textItem( t, 3 ) ) == entry->tStart ) &&
( atoi( textItem( t, 4 ) ) == entry->tEnd ) &&
entry->misMatches==0 && (entry->matches+entry->nCount) == entry->qSize)
res = 1;
else
res = 0;
textDestroy( t );
return res;
}
static void printMrfAlignBlocks (SamEntry *e, int _strand)
{
char strand = '.';
int len, intronic;
int q, pos;
int i;
Texta tokens;
tokens = textFieldtokP (e->cigar, "MN");
if (_strand == R_FIRST) {
if (e->flags & S_QUERY_STRAND)
strand = '-';
else
strand = '+';
} else {
if (e->flags & S_MATE_STRAND)
strand = '-';
else
strand = '+';
}
// Process first item in cigar
len = atoi (textItem (tokens, 0));
pos = e->pos;
q = 1;
printf ("%s:%c:%d:%d:%d:%d",
e->rname, strand, e->pos, pos + len - 1, 1, len);
pos += len - 1;
q += len;
// Process rest of cigar
if (arrayMax (tokens) > 2) {
for (i = 2; i < arrayMax (tokens) - 1; i += 2) {
len = atoi (textItem (tokens, i));
intronic = atoi (textItem (tokens, i - 1));
pos += intronic + 1;
printf(",%s:%c:%d:%d:%d:%d",
e->rname, strand, pos, pos + len - 1, q, q + len - 1);
pos += len - 1;
q += len;
}
}
textDestroy (tokens);
}
SEXP c_read_biokit_exprs (SEXP filename) {
LineStream ls;
char* line;
const int MAND_NCOL=7; // the first column is the row name, and column 2-7 are mandatory
int add_ncol=0;
Texta it;
Texta rnames=textCreate(128);
Array mrpkms=arrayCreate(128, double);
Array mreads=arrayCreate(128, int);
Array srpkms=arrayCreate(128, double);
Array sreads=arrayCreate(128, int);
Array mprop=arrayCreate(128, double);
Array allmap = arrayCreate(128, int);
Array annos=arrayCreate(128, Texta);
Texta anno=NULL; // must have a NULL assigned; otherwise textCreateClear leads to memory error
Stringa str=stringCreate(8);
SEXP R_rnames, R_mrpkms, R_mreads, R_srpkms, R_sreads, R_mprop, R_allmap, R_res;
SEXP R_colnames, R_class;
int nprot=0;
int i=0;
int j=0;
int nrow=0;
const char* fn=CHAR(STRING_ELT(filename, 0));
ls = ls_createFromFile(strdup(fn));
ls_nextLine(ls); // skip the first header line
while(line = ls_nextLine(ls)) {
it = textFieldtokP(line, "\t");
if(arrayMax(it)<MAND_NCOL)
error("Input file must contain no less than %d columns", MAND_NCOL);
textAdd(rnames, textItem(it, 0));
array(mrpkms, arrayMax(mrpkms), double)=atof(textItem(it, 1));
array(mreads, arrayMax(mreads), int)=atoi(textItem(it, 2));
array(srpkms, arrayMax(srpkms), double)=atof(textItem(it, 3));
array(sreads, arrayMax(sreads), int)=atoi(textItem(it, 4));
array(mprop, arrayMax(mprop), double)=atof(textItem(it, 5));
array(allmap, arrayMax(allmap), int)=atoi(textItem(it, 6));
add_ncol = max(arrayMax(it)-MAND_NCOL, add_ncol);
textCreateClear(anno, arrayMax(it)-MAND_NCOL);
for(i=MAND_NCOL; i<arrayMax(it); ++i) {
textAdd(anno, textItem(it, i));
}
array(annos, arrayMax(annos), Texta)=textClone(anno);
nrow++;
}
R_rnames=PROTECT(allocVector(STRSXP, nrow)); nprot++;
R_mrpkms=PROTECT(allocVector(REALSXP, nrow)); nprot++;
R_mreads=PROTECT(allocVector(INTSXP, nrow)); nprot++;
R_srpkms=PROTECT(allocVector(REALSXP, nrow)); nprot++;
R_sreads=PROTECT(allocVector(INTSXP, nrow)); nprot++;
R_mprop=PROTECT(allocVector(REALSXP, nrow)); nprot++;
R_allmap=PROTECT(allocVector(INTSXP, nrow)); nprot++;
for(i=0; i<nrow; ++i) {
SET_STRING_ELT(R_rnames, i, mkChar(textItem(rnames, i)));
REAL(R_mrpkms)[i]=arru(mrpkms, i, double);
INTEGER(R_mreads)[i]=arru(mreads, i, int);
REAL(R_srpkms)[i]=arru(srpkms, i, double);
INTEGER(R_sreads)[i]=arru(sreads, i, int);
REAL(R_mprop)[i]=arru(mprop, i, double);
INTEGER(R_allmap)[i]=arru(allmap, i, int);
}
R_res=PROTECT(allocVector(VECSXP, MAND_NCOL+add_ncol-1)); nprot++;
SET_VECTOR_ELT(R_res, 0, R_mrpkms);
SET_VECTOR_ELT(R_res, 1, R_mreads);
SET_VECTOR_ELT(R_res, 2, R_srpkms);
SET_VECTOR_ELT(R_res, 3, R_sreads);
SET_VECTOR_ELT(R_res, 4, R_mprop);
SET_VECTOR_ELT(R_res, 5, R_allmap);
for(i=0; i<add_ncol; ++i) {
SEXP R_anno=NULL;
R_anno=PROTECT(allocVector(STRSXP, nrow));
for(j=0; j<nrow; ++j) {
anno=array(annos, j, Texta);
if(arrayMax(anno)>i) {
SET_STRING_ELT(R_anno, j, mkChar(textItem(anno, i)));
} else {
SET_STRING_ELT(R_anno, j, R_NaString);
}
}
SET_VECTOR_ELT(R_res, i+MAND_NCOL-1, R_anno); // -1 because the first column is row name
UNPROTECT(1);
}
PROTECT(R_colnames=allocVector(STRSXP, MAND_NCOL+add_ncol-1)); nprot++;
PROTECT(R_class=allocVector(STRSXP, 1)); nprot++;
SET_STRING_ELT(R_colnames, 0, mkChar("RPKM_MultiMap"));
SET_STRING_ELT(R_colnames, 1, mkChar("ReadCount_MultiMap"));
SET_STRING_ELT(R_colnames, 2, mkChar("RPKM_UniqMap"));
SET_STRING_ELT(R_colnames, 3, mkChar("ReadCount_UniqMap"));
SET_STRING_ELT(R_colnames, 4, mkChar("MultiProp"));
SET_STRING_ELT(R_colnames, 5, mkChar("AllMappingReads"));
for(i=0; i<add_ncol; ++i) {
stringPrintf(str, "Annotation%d", i+1);
SET_STRING_ELT(R_colnames, i+MAND_NCOL-1,
mkChar(string(str)));
}
SET_STRING_ELT(R_class, 0, mkChar("data.frame"));
setAttrib(R_res, install("names"), R_colnames);
setAttrib(R_res, install("row.names"), R_rnames);
setAttrib(R_res, install("class"), R_class);
for(i=0; i<nrow; ++i) {
textDestroy(array(annos, i, Texta));
}
arrayDestroy(annos);
arrayDestroy(rnames);
arrayDestroy(mrpkms);
arrayDestroy(mreads);
arrayDestroy(srpkms);
arrayDestroy(sreads);
arrayDestroy(mprop);
arrayDestroy(allmap);
stringDestroy(str);
ls_destroy(ls);
UNPROTECT(nprot);
return(R_res);
}
int main (int argc, char **argv)
{
LineStream ls;
Texta tokens = NULL;
char *line;
int hasQual = 0;
int hasSeqs = 0;
int start=1;
ls = ls_createFromFile ("-");
while (line = ls_nextLine (ls)) {
// Put all the lines of the SAM header in comments
if (line[0] == '@') {
printf ("# %s\n", line);
continue;
}
// Parse each SAM entry and store into array
tokens = textFieldtokP (line, "\t");
if (arrayMax (tokens) < 11) {
textDestroy( tokens );
ls_destroy (ls);
die ("Invalid SAM entry: %s", line);
}
SamEntry *currSamE = NULL;
SamEntry *mateSamE = NULL;
AllocVar(currSamE );
int ret = generateSamEntry( tokens, currSamE, &hasSeqs, &hasQual );
textDestroy( tokens );
if ( ret==0 ) {
if ( isPaired ( currSamE ) )
ls_nextLine( ls ); // discarding next entry too (the mate)
destroySamEntry( currSamE );
freeMem( currSamE );
continue;
}
if ( isPaired( currSamE ) ) {
int hasQual2, hasSeq2;
AllocVar( mateSamE );
Texta secondEnd = NULL;
secondEnd = textFieldtok (ls_nextLine( ls ) , "\t");
ret = generateSamEntry( secondEnd, mateSamE, &hasSeq2, &hasQual2 );
textDestroy( secondEnd );
if( ret == 0 ) {
destroySamEntry( currSamE );
destroySamEntry( mateSamE );
freeMem( currSamE );
freeMem( mateSamE );
continue;
}
if (strcmp (currSamE->qname, mateSamE->qname) != 0) {
die ("Please note that for paired-end data, sam2mrf requires the mate pairs to be on subsequent lines. You may want to sort the SAM file first.\nEx: sort -r file.sam | sam2mrf > file.mrf\n");
}
}
// Print MRF headers
if( start ) {
printf ("%s", MRF_COLUMN_NAME_BLOCKS);
if (hasSeqs) printf("\t%s", MRF_COLUMN_NAME_SEQUENCE);
if (hasQual) printf("\t%s", MRF_COLUMN_NAME_QUALITY_SCORES);
printf ("\t%s\n", MRF_COLUMN_NAME_QUERY_ID);
start=0;
}
// Print AlignmentBlocks
printMrfAlignBlocks (currSamE, R_FIRST);
if( isPaired ( currSamE ) ) {
printf ("|");
printMrfAlignBlocks (mateSamE, R_SECOND);
}
seq_init();
// Print Sequence
if (hasSeqs) {
if (!currSamE->seq)
die ("Entry missing sequence column\n");
if( currSamE->flags & S_QUERY_STRAND )
seq_reverseComplement( currSamE->seq, strlen(currSamE->seq));
printf ("\t%s", currSamE->seq);
if (mateSamE) {
if (!mateSamE->seq)
die ("Entry missing sequence column\n");
if( mateSamE->flags & S_MATE_STRAND )
seq_reverseComplement( mateSamE->seq, strlen(mateSamE->seq));
printf ("|%s", mateSamE->seq);
}
}
// Print quality scores
if (hasQual) {
if (!currSamE->qual)
die ("Entry missing quality scores column\n");
printf ("\t%s", currSamE->qual);
if (mateSamE) {
if (!mateSamE->qual)
die ("Entry missing quality scores column\n");
printf ("|%s", mateSamE->qual);
}
}
// Print queryID
if (mateSamE) {
printf ("\t%s|%s", currSamE->qname,"2"); // No need to print out both IDs, but need the pipe symbol for consistency
}
else {
printf ("\t%s", currSamE->qname);
}
printf("\n");
destroySamEntry( currSamE );
freeMem( currSamE );
if( isPaired( currSamE ) ) {
destroySamEntry ( mateSamE );
freeMem( mateSamE );
}
}
// clean up
ls_destroy (ls);
return EXIT_SUCCESS;
}
void incl_getExonHlightFile (FILE *fp, Array regions, char *sdata_dir)
{
LineStream src;
FILE *out;
char *line;
Texta entry;
int i, astart, aend;
Stringa buffer = stringCreate (50);
stringPrintf (buffer, "%s/tmp/exons.hlight_s.txt", sdata_dir);
if (!(out = fopen (string (buffer), "w"))) {
fprintf (stderr, "Cannot open exons.hlight_s.txt\n");
return;
}
SRegion_t *tmp;
tmp = arrayp (regions, 0, SRegion_t);
if (tmp->chromosome == 0) {
fprintf (fp, "file = %s/exons.hlight.txt\n", sdata_dir);
}
else {
for (i = 0; i < arrayMax (regions); i++) {
tmp = arrayp (regions, i, SRegion_t);
if (tmp->chromosome == 23) {
stringPrintf (buffer, "%s/X/exons.hlight.txt", sdata_dir);
}
else if (tmp->chromosome == 24) {
stringPrintf (buffer, "%s/Y/exons.hlight.txt", sdata_dir);
}
else {
stringPrintf (buffer, "%s/%i/exons.hlight.txt", sdata_dir, tmp->chromosome);
}
if ((src = ls_createFromFile (string (buffer))) == NULL) {
fprintf (stderr, "Cannot open exons.hlight.txt\n");
return;
}
while ((line = ls_nextLine (src)) != NULL) {
entry = textFieldtokP (line, " ");
astart = atoi (textItem (entry, 1));
aend = atoi (textItem (entry, 2));
if ((astart >= tmp->start && astart <= tmp->end) ||
(aend >= tmp->start && aend <= tmp->end)) {
fprintf (out, "%s\n", line);
}
textDestroy (entry);
}
}
fprintf (fp, "file = %s/tmp/exons.hlight_s.txt\n", sdata_dir);
}
stringDestroy (buffer);
ls_destroy (src);
fclose (out);
}
