U _@sddlZddlZddlZddlZddlZddlZddlmZddlmZddlm Z ddlm Z ddl m Z ddlmZdd ZeZd d ZGd d d eZdS)N)bins)helpers) constantsmerge_criteria)Feature)FeatureNotFoundErrorcCs|d|jd<|S)a Helper for add_relation() Sets 'Parent' attribute to parent['ID'] Parameters ---------- parent : Feature Parent Feature :param parent: parent Feature child : Feature Child Feature Returns ------- Child Feature IDZParent) attributes)parentchildr 1lib/python3.8/site-packages/gffutils/interface.py assign_childsrcCs8t|dkr.dtdd|D|_||_nt|_|S)a Helper for FeatureDB.merge() to update source and assign children property Parameters ---------- feature : Feature feature to finalise feature_children list of children to assign Returns ------- feature, modified ,css|] }|jVqdSN)source).0r r r r <sz"_finalize_merge..)lenjoinsetrchildren no_childrenfeaturefeature_childrenr r r_finalize_merge+s  rc@seZdZdZddejdejfddZddZ dd Z d d Z d d Z ddZ dKddZdLddZdMddZdNddZddZdOddZdPddZdQd d!Zd"d#Zd$d%Zd&d'ZdRd(d)ZdSd+d,ZdTd-d.ZdUd/d0ZdVd1d2Zd3d4Zd5d6ZdWd9d:Z dXd;d<Z!e"j#e"j$e"j%e"j&fdfd=d>Z'd?e"j#e"j$e"j%e"j&fd@dfdAdBZ(dYdCdDZ)d7gdEgddFdfdGdHZ*ej+j,ej+dIe_+ej+j,ej+dIe_+eeeefD]Z-e-j+j,edJe-_+qdS)Z FeatureDBa5 limit : string or tuple Limit the results to a genomic region. If string, then of the form "seqid:start-end"; if tuple, then (seqid, start, end). strand : "-" | "+" | "." Limit the results to one strand featuretype : string or tuple Limit the results to one or several featuretypes. order_by : string or tuple Order results by one or many fields; the string or tuple items must be in: 'seqid', 'source', 'featuretype', 'start', 'end', 'score', 'strand', 'frame', 'attributes', 'extra'. reverse : bool Change sort order; only relevant if `order_by` is not None. By default, results will be in ascending order, so use `reverse=True` for descending. completely_within : bool If False (default), a single bp overlap with `limit` is sufficient to return a feature; if True, then the feature must be completely within `limit`. Only relevant when `limit` is not None. zutf-8Fc Cs:ddlm}t||jr*|j|_|j|_nXt|tjrD||_||_n>|dkrVtdn,t j |sntd|||_t |j|_|dk r||j_ tj|j_||_||_||_|j}|d|\} } | |_t| |_|dd d |D|_|d tt||_||| s6t!"d |jdS) aH Connect to a database created by :func:`gffutils.create_db`. Parameters ---------- dbfn : str Path to a database created by :func:`gffutils.create_db`. text_factory : callable Optionally set the way sqlite3 handles strings. Default is sqlite3.OptimizedUnicode, which returns ascii when possible, unicode otherwise encoding : str When non-ASCII characters are encountered, assume they are in this encoding. keep_order : bool If True, all features returned from this instance will have the order of their attributes maintained. This can be turned on or off database-wide by setting the `keep_order` attribute or with this kwarg, or on a feature-by-feature basis by setting the `keep_order` attribute of an individual feature. Default is False, since this includes a sorting step that can get time-consuming for many features. pragmas : dict Dictionary of pragmas to use when connecting to the database. See http://www.sqlite.org/pragma.html for the full list of possibilities, and constants.default_pragmas for the defaults. These can be changed later using the :meth:`FeatureDB.set_pragmas` method. Notes ----- `dbfn` can also be a subclass of :class:`_DBCreator`, useful for when :func:`gffutils.create_db` is provided the ``dbfn=":memory:"`` kwarg. rcreatez:memory:z:cannot connect to memory db; please provide the connectionzDatabase file %s does not existNz; SELECT version, dialect FROM meta z: SELECT directive FROM directives cSsg|]}|r|dqS)rr )rZ directiver r r sz&FeatureDB.__init__..z< SELECT base, n FROM autoincrements a"It appears that this database has not had the ANALYZE sqlite3 command run on it. Doing so can dramatically speed up queries, and is done by default for databases created with gffutils >0.8.7.1 (this database was created with version %s) Consider calling the analyze() method of this object.)#gffutilsr! isinstanceZ _DBCreatorconndbfnsqlite3Z Connection ValueErrorospathexistsZconnect text_factoryZRowZ row_factorydefault_encoding keep_ordersort_attribute_valuescursorexecutefetchoneversionrZ _unjsonifydialectZ directives collections defaultdictint_autoincrements set_pragmas _analyzedwarningswarn) selfr&r-r.pragmasr/r,r!cr3r4r r r__init__`sV7            zFeatureDB.__init__cCs>||_|j}|ddd|jD|jdS)a Set pragmas for the current database connection. Parameters ---------- pragmas : dict Dictionary of pragmas; see constants.default_pragmas for a template and http://www.sqlite.org/pragma.html for a full list. z; cSsg|] }d|qS)z PRAGMA %s=%sr rir r rr"sz)FeatureDB.set_pragmas..N)r>r%r0Z executescriptritemscommit)r=r>r?r r rr9s  zFeatureDB.set_pragmascKs4|d|j|d|j|d|jtf|S)zQ Returns a feature, adding additional database-specific defaults r4r.r/) setdefaultr4r.r/r)r=kwargsr r r_feature_returnerszFeatureDB._feature_returnercCs|d}tt|dkS)Nzp SELECT name FROM sqlite_master WHERE type='table' AND name='sqlite_stat1'; r)r1rlist)r=resr r rr:szFeatureDB._analyzedcCs@|j}|dg}|D]\}|dk r||qd|S)z: Returns the database schema as a string. z7 SELECT sql FROM sqlite_master N )r%r0r1appendr)r=r?resultsrBr r rschemas   zFeatureDB.schemac Cst|tr|j}|j}z|tjd|fWn2tj k rd|tjd| |j fYnX| }|dkr~t ||jf|S)Nz WHERE id = ?)r$ridr%r0r1r_SELECTr'ProgrammingErrordecoder-r2rrG)r=keyr?rLr r r __getitem__ s   zFeatureDB.__getitem__NcCsH|j}|dk r"|d|fn |d|}|dk rD|d}|S)a Simple count of features. Can be faster than "grep", and is faster than checking the length of results from :meth:`gffutils.FeatureDB.features_of_type`. Parameters ---------- featuretype : string Feature type (e.g., "gene") to count. If None, then count *all* features in the database. Returns ------- The number of features of this type, as an integer Nzd SELECT count() FROM features WHERE featuretype = ? z> SELECT count() FROM features r)r%r0r1r2)r= featuretyper?rLr r rcount_features_of_types z FeatureDB.count_features_of_typec cs@tjg||||||d\}}|||D]} |jf| Vq(dS)z Returns an iterator of :class:`gffutils.Feature` objects. Parameters ---------- {_method_doc} )argslimitrTorder_byreversestrandcompletely_withinNr make_query_executerG) r=rTrWrZrXrYr[queryrVrBr r rfeatures_of_typeAs  zFeatureDB.features_of_typegenec cs6|j|d}|D] }|gt||j}|VqdS)a For each parent of type `featuretype`, yield a list L of that parent and all of its children (`[parent] + list(children)`). The parent will always be L[0]. This is useful for "sanitizing" a GFF file for downstream tools. Additional kwargs are passed to :meth:`FeatureDB.children`, and will therefore only affect items L[1:] in each yielded list. )rTN) all_featuresrHrrN) r=rTlevelrXrYr[Z parent_recsZ parent_recZ unit_recordsr r riter_by_parent_childsYs  zFeatureDB.iter_by_parent_childsc cs@tjg||||||d\}}|||D]} |jf| Vq(dS)z Iterate through the entire database. Returns ------- A generator object that yields :class:`Feature` objects. Parameters ---------- {_method_doc} )rVrWrZrTrXrYr[Nr\) r=rWrZrTrXrYr[r_rVrBr r rrbos  zFeatureDB.all_featuresccs*|j}|d|D] \}|VqdS)z Iterate over feature types found in the database. Returns ------- A generator object that yields featuretypes (as strings) zC SELECT DISTINCT featuretype from features Nr%r0r1)r=r?rBr r r featuretypess   zFeatureDB.featuretypesc  cst|tr|j}djft} |g} d} |dk r>d} | |tj| | | |||| |d\} } | dd} | | | D]}|j f|VqtdS)a Parameters ---------- id : string or a Feature object level : None or int If `level=None` (default), then return all children regardless of level. If `level` is an integer, then constrain to just that level. {_method_doc} Returns ------- A generator object that yields :class:`Feature` objects. zs JOIN relations ON relations.{join_on} = features.id WHERE relations.{join_to} = ? Nzrelations.level = ?)rVotherextrarTrXrYrWr[ZSELECTzSELECT DISTINCT) r$rrNformatlocalsrKrr]replacer^rG)r=rNjoin_onjoin_torcrTrXrYr[rWrhrVZ level_clauser_rBr r r _relations.   zFeatureDB._relationc Cs|j|dd||||||d S)zP Return children of feature `id`. {_relation_docstring} r r rmrnrcrTrXrYrWr[ro)r=rNrcrTrXrYrWr[r r rrszFeatureDB.childrenc Cs|j|dd||||||d S)zO Return parents of feature `id`. {_relation_docstring} r r rprq)r=rNrcrTrXrYr[rWr r rparentsszFeatureDB.parentscCs*||_||_|j}||t||Sr) _last_query _last_argsr%r0r1tuple)r=r_rVr?r r rr^s  zFeatureDB._executecCs|j}||S)a Execute arbitrary queries on the db. .. seealso:: :class:`FeatureDB.schema` may be helpful when writing your own queries. Parameters ---------- query : str Query to execute -- trailing ";" optional. Returns ------- A sqlite3.Cursor object that can be iterated over. re)r=r_r?r r rr1s zFeatureDB.executecCs|d|jdS)zg Runs the sqlite ANALYZE command to potentially speed up queries dramatically. zANALYZE featuresN)r1r%rD)r=r r ranalyzes zFeatureDB.analyzeccs|dk r|dk s |dk s |dk r(tdt|tjr|d}t|dkr\|d}d\}}q|dd\}} t|dkr|d}| d \}}n6t|tr|j}|j}|j }|j }n|dd\}}}|rd } d } nd } d } ||}}g} g} |dk r | d| ||dk r6t |}| d| | ||dk r`t |}| d| | |d | } d}|dk r|dk r|r|tjkr|tjkrttj||dd}t|dkrd dd|D}d|}| |7} d tjd| |g}|dk rBt|tjr|g}d dd|D}|d|7}| ||dk rbd}||7}| ||j}||_| |_||||d|_||t| |D]}|jf|VqdS) a Return features within specified genomic coordinates. Specifying genomic coordinates can be done in a flexible manner Parameters ---------- region : string, tuple, or Feature instance If string, then of the form "seqid:start-end". If tuple, then (seqid, start, end). If :class:`Feature`, then use the features seqid, start, and end values. This argument is mutually exclusive with start/end/seqid. *Note*: By design, even if a feature is provided, its strand will be ignored. If you want to restrict the output by strand, use the separate `strand` kwarg. strand : + | - | . | None If `strand` is provided, then only those features exactly matching `strand` will be returned. So `strand='.'` will only return unstranded features. Default is `strand=None` which does not restrict by strand. seqid, start, end, strand Mutually exclusive with `region`. These kwargs can be used to approximate slice notation; see "Details" section below. featuretype : None, string, or iterable If not None, then restrict output. If string, then only report that feature type. If iterable, then report all featuretypes in the iterable. completely_within : bool By default (`completely_within=False`), returns features that partially or completely overlap `region`. If `completely_within=True`, features that are completely within `region` will be returned. Notes ------- The meaning of `seqid`, `start`, and `end` is interpreted as follows: ====== ====== ===== ====================================== seqid start end meaning ====== ====== ===== ====================================== str int int equivalent to `region` kwarg None int int features from all chroms within coords str None int equivalent to [:end] slice notation str int None equivalent to [start:] slice notation None None None equivalent to FeatureDB.all_features() ====== ====== ===== ====================================== If performance is a concern, use `completely_within=True`. This allows the query to be optimized by only looking for features that fall in the precise genomic bin (same strategy as UCSC Genome Browser and BEDTools). Otherwise all features' start/stop coords need to be searched to see if they partially overlap the region of interest. Examples -------- - `region(seqid="chr1", start=1000)` returns all features on chr1 that start or extend past position 1000 - `region(seqid="chr1", start=1000, completely_within=True)` returns all features on chr1 that start past position 1000. - `region("chr1:1-100", strand="+", completely_within=True)` returns only plus-strand features that completely fall within positions 1 to 100 on chr1. Returns ------- A generator object that yields :class:`Feature` objects. NzLIf region is supplied, do not supply seqid, start, or end as separate kwargs:rr)NN-z>=z<=<>z seqid = ?z start %s ?zend %s ?z AND rgFZoneiz or cSsg|]}dqS)zbin = ?r r_r r rr"sz$FeatureDB.region..z AND ( %s ) zWHERE cSsg|]}dqS)zfeaturetype = ?r r~r r rr"sz AND (%s) z and strand = ? )startendseqidregion)r(r$six string_typessplitrrrrrrZrKr7rrZMAX_CHROM_SIZErHrrOextendr%r0rsrtZ_contextr1rurG)r=rrrrrZrTr[ZtoksZcoordsZstart_opZend_oprVZposition_clauseZ _bin_clauseZ_binsr_Zfeature_clauseZ strand_clauser?rBr r rr sP                      zFeatureDB.regionTc cs.t|D]\}}|dkr&|j} |} q|j} |dkrDd| j|jf}| j|jkrVd} n|j} | j|jkrn|} q| } | j}| d7} | d8} |rt| j|j}ni}|r| |t j | | dd}d}t |d|| | d| d||d }|dkrz |j }Wnt k rd}YnX|jf|V|j} |} qdS) a Construct new features representing the space between features. For example, if `features` is a list of exons, then this method will return the introns. If `features` is a list of genes, then this method will return the intergenic regions. Providing N features will return N - 1 new features. This method purposefully does *not* do any merging or sorting of coordinates, so you may want to use :meth:`FeatureDB.merge` first, or when selecting features use the `order_by` kwarg, e.g., `db.features_of_type('gene', order_by=('seqid', 'start'))`. Parameters ---------- features : iterable of :class:`feature.Feature` instances Sorted, merged iterable new_featuretype : string or None The new features will all be of this type, or, if None (default) then the featuretypes will be constructed from the neighboring features, e.g., `inter_exon_exon`. merge_attributes : bool If True, new features' attributes will be a merge of the neighboring features' attributes. This is useful if you have provided a list of exons; the introns will then retain the transcript and/or gene parents as a single item. Otherwise, if False, the attribute will be a comma-separated list of values, potentially listing the same gene ID twice. attribute_func : callable or None If None, then nothing special is done to the attributes. If callable, then the callable accepts two attribute dictionaries and returns a single attribute dictionary. If `merge_attributes` is True, then `attribute_func` is called before `merge_attributes`. This could be useful for manually managing IDs for the new features. update_attributes : dict After attributes have been modified and merged, this dictionary can be used to replace parts of the attributes dictionary. Returns ------- A generator that yields :class:`Feature` objects rNz inter_%s_%s.rTr}Zgffutils_derived) rrrTrrscorerZframer bin) enumeratestoprrTrZchromrmerge_attributesr updaterdictr4AttributeErrorrG)r=featuresnew_featuretyperr4Zattribute_funcZupdate_attributesrBfZinterfeature_startZ last_featureZinterfeature_stopZ new_strandrZrZnew_attributesZnew_bin_idfieldsr r r interfeaturessn3      zFeatureDB.interfeaturesc Ks|r&t|jtjr&t|j|jd|j}d}d}t|trJ| }t|tjr\|g}t|t rl|g}|D]:}t|tjr|}n|j }| ||f| |||fqp|j |S)a Delete features from database. features : str, iterable, FeatureDB instance If FeatureDB, all features will be used. If string, assume it's the ID of the feature to remove. Otherwise, assume it's an iterable of Feature objects. The classes in gffutils.iterators may be helpful in this case. make_backup : bool If True, and the database you're about to update is a file on disk, makes a copy of the existing database and saves it with a .bak extension. Returns ------- FeatureDB object, with features deleted. .bakz3 DELETE FROM features WHERE id = ? zE DELETE FROM relations WHERE parent = ? OR child = ? )r$r&rrshutilcopy2r%r0rrbrrNr1rD) r=r make_backuprFr?Zquery1Zquery2rrr r rdelete6s(      zFeatureDB.deletec KsRddlm}ddlm}|r>t|jtjr>t|j|jdi}| D]\}}|t j krJ|||<qJ|j |f|}|j |d<|jddkrd|krd d d |d<|jf||j|jd |} n@|jdd krd|krd|d<|jf||j|jd |} nt||jd\} |_t| dkr&| S| || | |j | j |S)a Update the on-disk database with features in `data`. WARNING: If you used any non-default kwargs for gffutils.create_db when creating the database in the first place (especially `disable_infer_transcripts` or `disable_infer_genes`) then you should use those same arguments here. The returned object is the same FeatureDB, but since it is pointing to the same database and that has been just updated, the new features can now be accessed. Parameters ---------- data : str, iterable, FeatureDB instance If FeatureDB, all data will be used. If string, assume it's a filename of a GFF or GTF file. Otherwise, assume it's an iterable of Feature objects. The classes in gffutils.iterators may be helpful in this case. make_backup : bool If True, and the database you're about to update is a file on disk, makes a copy of the existing database and saves it with a .bak extension. kwargs : Additional kwargs are passed to gffutils.create_db; see the help for that function for details. Returns ------- Returns self but with the underlying database updated. rr ) iteratorsrr8ZfmtZgtfZid_specZgene_idZ transcript_id)raZ transcript)datar&r4Zgff3r r)r#r!rr$r&rrrrrCr_iterator_kwargsZ DataIteratorr8r4Z _GTFDBCreatorZ _GFFDBCreatorr(peekZ_iterrZ_populate_from_linesZ_update_relationsZ _finalizer) r=rrrFr!rrkvZdbrr r rrdsT$       zFeatureDB.updatecCst|tjr||}t|tjr(||}|j}|d|j|j|f|dk rf|||}||||dk r|||}||||j|S)a Manually add relations to the database. Parameters ---------- parent : str or Feature instance Parent feature to add. child : str or Feature instance Child feature to add level : int Level of the relation. For example, if parent is a gene and child is an mRNA, then you might want level to be 1. But if child is an exon, then level would be 2. parent_func, child_func : callable These optional functions control how attributes are updated in the database. They both have the signature `func(parent, child)` and must return a [possibly modified] Feature instance. For example, we could add the child's database id as the "child" attribute in the parent:: def parent_func(parent, child): parent.attributes['child'] = child.id and add the parent's "gene_id" as the child's "Parent" attribute:: def child_func(parent, child): child.attributes['Parent'] = parent['gene_id'] Returns ------- FeatureDB object with new relations added. zb INSERT INTO relations (parent, child, level) VALUES (?, ?, ?)N) r$rrr%r0r1rN_updaterD)r=r r rcZ parent_func child_funcr?r r r add_relations %         zFeatureDB.add_relationcCs0t||jgg}|jtjft|dSr)rHastuplerNr1rZ_UPDATEru)r=rr0valuesr r rrs zFeatureDB._updatec CsHz|tj|Wn,tjk rB|tj||jYnXdS)z5 Insert a feature into the database. N)r1rZ_INSERTrr'rPr-)r=rr0r r r_inserts zFeatureDB._insertexonintronc #srsdkr dkr tdr4fdd}nrFfdd}|D]6}j|d|dd}j|||jd D] } | VqvqLdS) a Create introns from existing annotations. Parameters ---------- exon_featuretype : string Feature type to use in order to infer introns. Typically `"exon"`. grandparent_featuretype : string If `grandparent_featuretype` is not None, then group exons by children of this featuretype. If `granparent_featuretype` is "gene" (default), then introns will be created for all first-level children of genes. This may include mRNA, rRNA, ncRNA, etc. If you only want to infer introns from one of these featuretypes (e.g., mRNA), then use the `parent_featuretype` kwarg which is mutually exclusive with `grandparent_featuretype`. parent_featuretype : string If `parent_featuretype` is not None, then only use this featuretype to infer introns. Use this if you only want a subset of featuretypes to have introns (e.g., "mRNA" only, and not ncRNA or rRNA). Mutually exclusive with `grandparent_featuretype`. new_featuretype : string Feature type to use for the inferred introns; default is `"intron"`. merge_attributes : bool Whether or not to merge attributes from all exons. If False then no attributes will be created for the introns. Returns ------- A generator object that yields :class:`Feature` objects representing new introns Notes ----- The returned generator can be passed directly to the :meth:`FeatureDB.update` method to permanently add them to the database, e.g., :: db.update(db.create_introns()) NzSexactly one of `grandparent_featuretype` or `parent_featuretype` should be providedc3s.D]}j|ddD] }|Vqq dS)Nr)rc)r`r)rar )grandparent_featuretyper=r r child_gen4sz+FeatureDB.create_introns..child_genc3sD] }|Vq dSr)r`)r )parent_featuretyper=r rr9srr)rcrTrX)rrr4)r(rrr4) r=Zexon_featuretyperrrrrr exonsrr )rrr=rcreate_intronss*1   zFeatureDB.create_intronsc csFt|}t|dkrt|r"d}n.dd|D}tt|dkrHtd|d}dd|D}tt|dkrvtd|d}|dj}|dj}|dd D]`} | j|dkr| j|kr| j}qqqt| j d| j ||d|dd d } |j f| V| j}| j}qt|dkr|d} t| j d| j ||d|dd d } |j f| Vd S) ay DEPRECATED, only retained here for backwards compatibility. Please use merge(). Merge overlapping features together. Parameters ---------- features : iterator of Feature instances ignore_strand : bool If True, features on multiple strands will be merged, and the final strand will be set to '.'. Otherwise, ValueError will be raised if trying to merge features on differnt strands. Returns ------- A generator object that yields :class:`Feature` objects representing the newly merged features. rrcSsg|] }|jqSr )rZrAr r rr"gsz(FeatureDB._old_merge..rz?Specify ignore_strand=True to force merging of multiple strandscSsg|] }|jqSr )rrAr r rr"nsz,Merging multiple chromosomes not implementedNrg) rrrTrrrrZrr ) rHr StopIterationrr(NotImplementedErrorrrrrrTrG) r=r ignore_strandrZZstrandsZchromsrZcurrent_merged_startZcurrent_merged_stoprZmerged_featurer r r _old_mergeFs`      zFeatureDB._old_mergec#sBt|ts2z t|}Wntk r0|g}YnXt|}d}dg|D]މdkrtfdd|Dr|gqJttVd}qJtdkrtfdd|DrnttVd}qJtfdd|Drtdkrvt  d=d =d =d =d =|j f|sh|j j d7<j d t|j j }|d<|_jjdkrjdj7_jjkrd_jjkrd_j j krd_ jjkrj_jjkr*j_qJtVgd}qJr>tVdS)a Merge features matching criteria together Returned Features have a special property called 'children' that is a list of the component features. This only exists for the lifetime of the Feature instance. Parameters ---------- features : iterable Iterable of Feature instances to merge merge_criteria : list List of merge criteria callbacks. All must evaluate to True in order for a feature to be merged. See notes below on callback signature. multiline : bool True to emit multiple features with the same ID attribute, False otherwise. Returns ------- Generator yielding merged Features Notes ----- See the `gffutils.merge_criteria` module (imported here as `mc`) for existing callback functions. For writing custom callbacks, functions must have the following signature:: callback( acc: gffutils.Feature, cur: gffutils.Feature, components: [gffutils.Feature] ) -> bool Where: - `acc`: current accumulated feature - `cur`: candidate feature to merge - `components`: list of features that compose acc The function should return True to merge `cur` into `acc`, False to set `cur` to `acc` (that is, start a new merged feature). If merge criteria allows different feature types then the merged features' feature types should have their featuretype property reassigned to a more specific ontology value. Nc3s|]}|VqdSrr rZcriteriarr rrsz"FeatureDB.merge..rc3s|]}|VqdSrr r)current_mergedrr rrsc3s|]}|VqdSrr rrrrr rrsrr rir4r.r/rr rrZsequence_feature)r$rH TypeErroriterallrrrrKvarscopyrGr8rTstrrNrrrZrrr)r=rrZ multilineZlast_idr rrmergesv8                zFeatureDB.merge)rrTrZrrc Cst|s d}g}|D]v}|j|j||d|dD]X}|jr0|||j|r^||jn|jD]}|j||dt dqd| |q0q0q|S)a Merge all features in database according to criteria. Merged features will be assigned as children of the merged record. The resulting records are added to the database. Parameters ---------- merge_order : list Ordered list of columns with which to group features before evaluating criteria merge_criteria : list List of merge criteria callbacks. See merge(). featuretypes_groups : list iterable of sets of featuretypes to merge together exclude_components : bool True: child features will be discarded. False to keep them. Returns ------- list of merge features rrTrXrr)r) rrrbrrr%r0rrrrK) r=Z merge_orderrZfeaturetypes_groupsZexclude_componentsZresult_featuresZ featuregroupZmergedr r r r merge_all3s    zFeatureDB.merge_allcCs@|j||dd}|r"|j||d}d}|D]}|t|7}q*|S)a Total bp of all children of a featuretype. Useful for getting the exonic bp of an mRNA. Parameters ---------- feature : str or Feature instance child_featuretype : str Which featuretype to consider. For example, to get exonic bp of an mRNA, use `child_featuretype='exon'`. merge : bool Whether or not to merge child features together before summing them. ignore_strand : bool If True, then overlapping features on different strands will be merged together; otherwise, merging features with different strands will result in a ValueError. Returns ------- Integer representing the total number of bp. rr)rr)rrr)r=rZchild_featuretyperrrtotalr r r r children_bpgszFeatureDB.children_bpZCDSr csL|r|rtdt|j||dd}t|dkr6|g}||}|dj}|dj} ||jkrntd||jf| |jkrtd| |jf|dkrd }|d d }|j} |jd |j} t j } d t _ z||d} Wnt k rd} YnX| t _ |j }|dkr d}|j }|}t|}dd|D}fdd|D}|rt|j||dd}t|dkrt|j}|j}n|djd }|dj}|rt|j||dd}t|dkr|j}|j}n|dj}|djd }|d|d}|| kstd|| f| | | ||||||dtt|dtt|g }dtt|S)a; Converts `feature` into a BED12 format. GFF and GTF files do not necessarily define genes consistently, so this method provides flexiblity in specifying what to call a "transcript". Parameters ---------- feature : str or Feature instance In most cases, this feature should be a transcript rather than a gene. block_featuretype : str or list Which featuretype to use as the exons. These are represented as blocks in the BED12 format. Typically 'exon'. Use the `thick_featuretype` and `thin_featuretype` arguments to control the display of CDS as thicker blocks and UTRs as thinner blocks. Note that the features for `thick` or `thin` are *not* automatically included in the blocks; if you do want them included, then those featuretypes should be added to this `block_features` list. If no child features of type `block_featuretype` are found, then the full `feature` is returned in BED12 format as if it had a single exon. thick_featuretype : str or list Child featuretype(s) to use in order to determine the boundaries of the "thick" blocks. In BED12 format, these represent coding sequences; typically this would be set to "CDS". This argument is mutually exclusive with `thin_featuretype`. Specifically, the BED12 thickStart will be the start coord of the first `thick` item and the thickEnd will be the stop coord of the last `thick` item. thin_featuretype : str or list Child featuretype(s) to use in order to determine the boundaries of the "thin" blocks. In BED12 format, these represent untranslated regions. Typically "utr" or ['three_prime_UTR', 'five_prime_UTR']. Mutually exclusive with `thick_featuretype`. Specifically, the BED12 thickStart field will be the stop coord of the first `thin` item and the thickEnd field will be the start coord of the last `thin` item. name_field : str Which attribute of `feature` to use as the feature's name. If this field is not present, a "." placeholder will be used instead. color : None or str If None, then use black (0,0,0) as the RGB color; otherwise this should be a comma-separated string of R,G,B values each of which are integers in the range 0-255. zACan only specify one of `thick_featuertype` or `thin_featuretype`rrrz=Start of first exon (%s) does not match start of feature (%s)z8End of last exon (%s) does not match end of feature (%s)Nz0,0,0rrgrTr0cSsg|] }t|qSr )rrAr r rr"sz#FeatureDB.bed12..csg|]}|jdqS)r)rrAZ chromStartr rr"sztst=%s; chromEnd=%sr )r(rHrrrrrlstriprrZalways_return_listKeyErrorrrZAssertionErrorrmapr)r=rZblock_featuretypeZthick_featuretypeZthin_featuretypeZ name_fieldZcolorrfirstZlastrZchromEndZorignamerrZZitemRgbZ blockCountZ blockSizesZ blockStartsZthickZ thickStartZthickEndZthinZtstrr rrbed12s=              zFeatureDB.bed12)Z_relation_docstring) _method_doc)N)NNNFF)raNNFF)NNNNFF)NNNFFN)NNNFNF)NNNFFN)NNNNNNF)NTNNN)T)T)NN)rraNrT)F)rFF).__name__ __module__ __qualname__rrZdefault_pragmasr'ZOptimizedUnicoder@r9rGr:rMrSrUr`rdrbrfrorrrr^r1rvrrrrrrrrrmcrZoverlap_end_inclusiverZZ feature_typerrrr__doc__rjmethodr r r rrBs x  &    8   4 w . P 9  J d   4 ( r)r5r)rr'rr;r#rrrrrZgffutils.featurerZgffutils.exceptionsrrrurrobjectrr r r rs