obitools/obitools3
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 66 Wiki Activity

Taxonomy: new functions to find taxa by name

Browse Source
This commit is contained in:
Celine Mercier
2020-10-30 10:45:20 +01:00
parent b9b4cec5b5
commit 112e12cab0
7 changed files with 174 additions and 58 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
python/obitools3/dms/capi/obitaxonomy.pxd
View File

@ -7,6 +7,8 @@ from libc.stdint cimport int32_t
cdef extern from "obidms_taxonomy.h" nogil:
extern int MIN_LOCAL_TAXID
struct ecotxnode :
int32_t taxid
int32_t rank
@ -18,6 +20,13 @@ cdef extern from "obidms_taxonomy.h" nogil:
ctypedef ecotxnode ecotx_t
struct econame_t : # can't get this struct to be accepted by Cython ('unknown size')
char* name
char* class_name
int32_t is_scientific_name
ecotxnode* taxon
struct ecotxidx_t :
int32_t count
int32_t max_taxid
@ -30,9 +39,14 @@ cdef extern from "obidms_taxonomy.h" nogil:
char** label
struct econameidx_t :
int32_t count
econame_t* names
struct OBIDMS_taxonomy_t :
ecorankidx_t* ranks
# econameidx_t* names
econameidx_t* names
ecotxidx_t* taxa
ctypedef OBIDMS_taxonomy_t* OBIDMS_taxonomy_p
@ -51,7 +65,11 @@ cdef extern from "obidms_taxonomy.h" nogil:
ecotx_t* obi_taxo_get_parent_at_rank(ecotx_t* taxon, int32_t rankidx)
ecotx_t* obi_taxo_get_taxon_with_taxid(OBIDMS_taxonomy_p taxonomy, int32_t taxid)
char* obi_taxo_get_name_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx)
ecotx_t* obi_taxo_get_taxon_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx)
bint obi_taxo_is_taxon_under_taxid(ecotx_t* taxon, int32_t other_taxid)
ecotx_t* obi_taxo_get_species(ecotx_t* taxon, OBIDMS_taxonomy_p taxonomy)
@ -71,4 +89,4 @@ cdef extern from "obidms_taxonomy.h" nogil:
int obi_taxo_add_preferred_name_with_taxon(OBIDMS_taxonomy_p tax, ecotx_t* taxon, const char* preferred_name)
const char* obi_taxo_rank_index_to_label(int32_t rank_idx, ecorankidx_t* ranks)

1
python/obitools3/dms/capi/obiview.pxd
View File

@ -27,6 +27,7 @@ cdef extern from "obiview.h" nogil:
extern const_char_p REVERSE_QUALITY_COLUMN
extern const_char_p REVERSE_SEQUENCE_COLUMN
extern const_char_p COUNT_COLUMN
extern const_char_p SCIENTIFIC_NAME_COLUMN
extern const_char_p TAXID_COLUMN
extern const_char_p MERGED_TAXID_COLUMN
extern const_char_p MERGED_PREFIX

3
python/obitools3/dms/taxo/taxo.pxd
View File

@ -11,11 +11,14 @@ cdef class Taxonomy(OBIWrapper) :
cdef bytes _name
cdef DMS _dms
cdef list _ranks
cdef dict _name_dict
cdef inline OBIDMS_taxonomy_p pointer(self)
cdef fill_name_dict(self)
cpdef Taxon get_taxon_by_idx(self, int idx)
cpdef Taxon get_taxon_by_taxid(self, int taxid)
cpdef Taxon get_taxon_by_name(self, object taxon_name, object restricting_taxid=*)
cpdef write(self, object prefix)
cpdef int add_taxon(self, str name, str rank_name, int parent_taxid, int min_taxid=*)
cpdef object get_species(self, int taxid)

53
python/obitools3/dms/taxo/taxo.pyx
View File

@ -15,7 +15,11 @@ from ..capi.obitaxonomy cimport obi_taxonomy_exists, \
obi_taxo_get_species, \
obi_taxo_get_genus, \
obi_taxo_get_family, \
ecotx_t
ecotx_t, \
econame_t, \
obi_taxo_get_name_from_name_idx, \
obi_taxo_get_taxon_from_name_idx
from cpython.pycapsule cimport PyCapsule_New, PyCapsule_GetPointer
import tarfile
@ -24,11 +28,29 @@ from libc.stdlib cimport free
cdef class Taxonomy(OBIWrapper) :
# TODO function to import taxonomy?
# TODO function to import taxonomy?
cdef inline OBIDMS_taxonomy_p pointer(self) :
return <OBIDMS_taxonomy_p>(self._pointer)
cdef fill_name_dict(self):
print("Indexing taxon names...")
cdef OBIDMS_taxonomy_p pointer = self.pointer()
cdef ecotx_t* taxon_p
cdef object taxon_capsule
cdef bytes name
cdef int count
cdef int n
count = (<OBIDMS_taxonomy_p>pointer).names.count
for n in range(count) :
name = obi_taxo_get_name_from_name_idx(pointer, n)
taxon_p = obi_taxo_get_taxon_from_name_idx(pointer, n)
taxon_capsule = PyCapsule_New(taxon_p, NULL, NULL)
self._name_dict[name] = Taxon(taxon_capsule, self)
@staticmethod
def exists(DMS dms, object name) :
@ -75,7 +97,8 @@ cdef class Taxonomy(OBIWrapper) :
taxo._dms = dms
taxo._name = tobytes(name)
taxo._name_dict = {}
taxo.fill_name_dict()
taxo._ranks = []
for r in range((<OBIDMS_taxonomy_p>pointer).ranks.count) :
taxo._ranks.append(obi_taxo_rank_index_to_label(r, (<OBIDMS_taxonomy_p>pointer).ranks))
@ -118,7 +141,8 @@ cdef class Taxonomy(OBIWrapper) :
taxo._dms = dms
taxo._name = folder_path
taxo._name_dict = {}
taxo.fill_name_dict()
taxo._ranks = []
for r in range((<OBIDMS_taxonomy_p>pointer).ranks.count) :
taxo._ranks.append(obi_taxo_rank_index_to_label(r, (<OBIDMS_taxonomy_p>pointer).ranks))
@ -129,8 +153,8 @@ cdef class Taxonomy(OBIWrapper) :
def __getitem__(self, object ref):
if type(ref) == int :
return self.get_taxon_by_taxid(ref)
else :
raise NotImplementedError()
elif type(ref) == str or type(ref) == bytes :
return self.get_taxon_by_name(ref)
cpdef Taxon get_taxon_by_taxid(self, int taxid):
@ -143,6 +167,19 @@ cdef class Taxonomy(OBIWrapper) :
return Taxon(taxon_capsule, self)
cpdef Taxon get_taxon_by_name(self, object taxon_name, object restricting_taxid=None):
taxon = self._name_dict.get(tobytes(taxon_name), None)
if not taxon:
return None
elif restricting_taxid:
if self.is_ancestor(restricting_taxid, taxon.taxid):
return taxon
else:
return None
else:
return taxon
cpdef Taxon get_taxon_by_idx(self, int idx):
cdef ecotx_t* taxa
cdef ecotx_t* taxon_p
@ -232,7 +269,7 @@ cdef class Taxonomy(OBIWrapper) :
taxa = self.pointer().taxa.taxon
# Yield each taxid
# Yield each taxon
for t in range(self.pointer().taxa.count):
taxon_p = <ecotx_t*> (taxa+t)
taxon_capsule = PyCapsule_New(taxon_p, NULL, NULL)

12
src/obidms_taxonomy.c
View File

@ -3649,6 +3649,18 @@ ecotx_t* obi_taxo_get_taxon_with_taxid(OBIDMS_taxonomy_p taxonomy, int32_t taxid
}
char* obi_taxo_get_name_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx)
{
return (((taxonomy->names)->names)[idx]).name;
}
ecotx_t* obi_taxo_get_taxon_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx)
{
return (((taxonomy->names)->names)[idx]).taxon;
}
int obi_taxo_is_taxon_under_taxid(ecotx_t* taxon, int32_t other_taxid) // TODO discuss that this doesn't work with deprecated taxids
{
ecotx_t* next_parent;

45
src/obidms_taxonomy.h
View File

@ -447,8 +447,51 @@ ecotx_t* obi_taxo_get_superkingdom(ecotx_t* taxon, OBIDMS_taxonomy_p taxonomy);
const char* obi_taxo_rank_index_to_label(int32_t rank_idx, ecorankidx_t* ranks);
// TODO
/**
* @brief Function checking whether a taxid is included in a subset of the taxonomy.
*
* @param taxonomy A pointer on the taxonomy structure.
* @param restrict_to_taxids An array of taxids. The researched taxid must be under at least one of those array taxids.
* @param count Number of taxids in restrict_to_taxids.
* @param taxid The taxid to check.
*
* @returns A value indicating whether the taxid is included in the chosen subset of the taxonomy.
* @retval 0 if the taxid is not included in the subset of the taxonomy.
* @retval 1 if the taxid is included in the subset of the taxonomy.
*
* @since October 2020
* @author Celine Mercier (celine.mercier@metabarcoding.org)
*/
int obi_taxo_is_taxid_included(OBIDMS_taxonomy_p taxonomy,
int32_t* restrict_to_taxids,
int32_t count,
int32_t taxid);
/**
* @brief Function returning the name of a taxon from its index in the taxonomy name index (econameidx_t).
*
* @param taxonomy A pointer on the taxonomy structure.
* @param idx The index at which the name is in the taxonomy name index (econameidx_t).
*
* @returns The taxon name.
*
* @since October 2020
* @author Celine Mercier (celine.mercier@metabarcoding.org)
*/
char* obi_taxo_get_name_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx);
/**
* @brief Function returning a taxon structure from its index in the taxonomy name index (econameidx_t).
*
* @param taxonomy A pointer on the taxonomy structure.
* @param idx The index at which the taxon is in the taxonomy name index (econameidx_t).
*
* @returns The taxon structure.
*
* @since October 2020
* @author Celine Mercier (celine.mercier@metabarcoding.org)
*/
ecotx_t* obi_taxo_get_taxon_from_name_idx(OBIDMS_taxonomy_p taxonomy, int32_t idx);

96
src/obiview.h
View File

@ -30,54 +30,56 @@
#include "obiblob.h"
#define OBIVIEW_NAME_MAX_LENGTH (249) /**< The maximum length of an OBIDMS view name, without the extension.
*/
#define VIEW_TYPE_MAX_LENGTH (1024) /**< The maximum length of the type name of a view.
*/
#define LINES_COLUMN_NAME "LINES" /**< The name of the column containing the line selections
* in all views.
*/
#define VIEW_TYPE_NUC_SEQS "NUC_SEQS_VIEW" /**< The type name of views based on nucleotide sequences
* and their metadata.
*/
#define NUC_SEQUENCE_COLUMN "NUC_SEQ" /**< The name of the column containing the nucleotide sequences
* in NUC_SEQS_VIEW views.
*/
#define ID_COLUMN "ID" /**< The name of the column containing the sequence identifiers
* in NUC_SEQS_VIEW views.
*/
#define DEFINITION_COLUMN "DEFINITION" /**< The name of the column containing the sequence definitions
* in NUC_SEQS_VIEW views.
*/
#define QUALITY_COLUMN "QUALITY" /**< The name of the column containing the sequence qualities
* in NUC_SEQS_VIEW views.
*/
#define REVERSE_QUALITY_COLUMN "REVERSE_QUALITY" /**< The name of the column containing the sequence qualities
* of the reverse read (generated by ngsfilter, used by alignpairedend).
*/
#define OBIVIEW_NAME_MAX_LENGTH (249) /**< The maximum length of an OBIDMS view name, without the extension.
*/
#define VIEW_TYPE_MAX_LENGTH (1024) /**< The maximum length of the type name of a view.
*/
#define LINES_COLUMN_NAME "LINES" /**< The name of the column containing the line selections
* in all views.
*/
#define VIEW_TYPE_NUC_SEQS "NUC_SEQS_VIEW" /**< The type name of views based on nucleotide sequences
* and their metadata.
*/
#define NUC_SEQUENCE_COLUMN "NUC_SEQ" /**< The name of the column containing the nucleotide sequences
* in NUC_SEQS_VIEW views.
*/
#define ID_COLUMN "ID" /**< The name of the column containing the sequence identifiers
* in NUC_SEQS_VIEW views.
*/
#define DEFINITION_COLUMN "DEFINITION" /**< The name of the column containing the sequence definitions
* in NUC_SEQS_VIEW views.
*/
#define QUALITY_COLUMN "QUALITY" /**< The name of the column containing the sequence qualities
* in NUC_SEQS_VIEW views.
*/
#define REVERSE_QUALITY_COLUMN "REVERSE_QUALITY" /**< The name of the column containing the sequence qualities
* of the reverse read (generated by ngsfilter, used by alignpairedend).
*/
#define REVERSE_SEQUENCE_COLUMN "REVERSE_SEQUENCE" /**< The name of the column containing the sequence
* of the reverse read (generated by ngsfilter, used by alignpairedend).
*/
#define QUALITY_COLUMN "QUALITY" /**< The name of the column containing the sequence qualities
* in NUC_SEQS_VIEW views.
*/
#define COUNT_COLUMN "COUNT" /**< The name of the column containing the sequence counts
* in NUC_SEQS_VIEW views.
*/
#define TAXID_COLUMN "TAXID" /**< The name of the column containing the taxids. TODO subtype of INT column?
*/
#define MERGED_TAXID_COLUMN "MERGED_TAXID" /**< The name of the column containing the merged taxids information.
*/
#define MERGED_PREFIX "MERGED_" /**< The prefix to prepend to column names when merging informations during obi uniq.
*/
#define TAXID_DIST_COLUMN "TAXID_DIST" /**< The name of the column containing a dictionary of taxid:[list of ids] when merging informations during obi uniq.
*/
#define MERGED_COLUMN "MERGED" /**< The name of the column containing a list of ids when merging informations during obi uniq.
*/
#define ID_PREFIX "seq" /**< The default prefix of sequence identifiers in automatic ID columns.
*/
#define PREDICATE_KEY "predicates" /**< The key used in the json-formatted view comments to store predicates.
*/
* of the reverse read (generated by ngsfilter, used by alignpairedend).
*/
#define QUALITY_COLUMN "QUALITY" /**< The name of the column containing the sequence qualities
* in NUC_SEQS_VIEW views.
*/
#define COUNT_COLUMN "COUNT" /**< The name of the column containing the sequence counts
* in NUC_SEQS_VIEW views.
*/
#define SCIENTIFIC_NAME_COLUMN "SCIENTIFIC_NAME" /**< The name of the column containing the taxon scientific name.
*/
#define TAXID_COLUMN "TAXID" /**< The name of the column containing the taxids. TODO subtype of INT column?
*/
#define MERGED_TAXID_COLUMN "MERGED_TAXID" /**< The name of the column containing the merged taxids information.
*/
#define MERGED_PREFIX "MERGED_" /**< The prefix to prepend to column names when merging informations during obi uniq.
*/
#define TAXID_DIST_COLUMN "TAXID_DIST" /**< The name of the column containing a dictionary of taxid:[list of ids] when merging informations during obi uniq.
*/
#define MERGED_COLUMN "MERGED" /**< The name of the column containing a list of ids when merging informations during obi uniq.
*/
#define ID_PREFIX "seq" /**< The default prefix of sequence identifiers in automatic ID columns.
*/
#define PREDICATE_KEY "predicates" /**< The key used in the json-formatted view comments to store predicates.
*/
/**