B b@s\dZddlmZddlZddlmZGdddeZGdddeZd d d Z dd d Z dS)zCode to work with the BLAST XML output. The BLAST XML DTD file is on the NCBI FTP site at: ftp://ftp.ncbi.nlm.nih.gov/blast/documents/xml/NCBI_BlastOutput.dtd )RecordN)ContentHandlerc@s:eZdZdZdddZddZddZd d Zd d Zd S) _XMLparserzGeneric SAX Parser (PRIVATE). Just a very basic SAX parser. Redefine the methods startElement, characters and endElement. rcCs(g|_d|_||_g|_d|_d|_dS)zuInitialize the parser. Arguments: - debug - integer, amount of debug information to print N)_tag_value_debug_debug_ignore_list_method_name_level _method_map)selfdebugr0lib/python3.7/site-packages/Bio/Blast/NCBIXML.py__init__s z_XMLparser.__init__cCs|j|t|jdkr(||dSd||}||jkrd|j||jdkrtd|n,|jdkr||jkrtd||j||j rt d|j |fd |_ dS) zFound XML start tag. No real need of attr, BLAST DTD doesn't use them Arguments: - name -- name of the tag - attr -- tag attributes rNZstart_zNCBIXML: Parsed: zNCBIXML: Ignored: z,What should we do with %s before the %r tag?r) rappendlen _on_root_node_node_method_namer r printr rstrip ValueError)r nameattrmethodrrr startElement*s"          z_XMLparser.startElementcCs|j|7_dS)zOFound some text. Arguments: - ch -- characters read N)r)r Zchrrr charactersRsz_XMLparser.characterscCsd||}||jkrB|j||jdkrttd||jfn2|jdkrt||jkrttd||jf|j|d|_|jdS)zLFound XML end tag. Arguments: - name -- tag name Zend_zNCBIXML: Parsed: %s %srzNCBIXML: Ignored: %s %srN) rr r rrr rrpop)r rrrrr endElement[s       z_XMLparser.endElementcCs&|jdkr|Sd|j|j dS)Nr/)r joinr)r rrrrrxs z_XMLparser._node_method_nameN)r) __name__ __module__ __qualname____doc__rrrr"rrrrrrs  ( rc@seZdZdZdddZddZddZd d Zd d Zd dZ ddZ ddZ ddZ ddZ ddZddZddZddZdd Zd!d"Zd#d$Zd%d&Zd'd(Zd)d*Zd+d,Zd-d.Zd/d0Zd1d2Zd3d4Zd5d6Zd7d8Zd9d:Zd;d<Z d=d>Z!d?d@Z"dAdBZ#dCdDZ$dEdFZ%dGdHZ&dIdJZ'dKdLZ(dMdNZ)dOdPZ*dQdRZ+dSdTZ,dUdVZ-dWdXZ.dYdZZ/d[d\Z0d]d^Z1d_d`Z2dadbZ3dcddZ4dedfZ5dgdhZ6didjZ7dkdlZ8dmdnZ9dodpZ:dqdrZ;dsdtZdydzZ?d{d|Z@d}d~ZAddZBddZCddZDdS) BlastParseraParse XML BLAST data into a Record.Blast object. Parses XML output from BLAST (direct use discouraged). This (now) returns a list of Blast records. Historically it returned a single Blast record. You are expected to use this via the parse or read functions. All XML 'action' methods are private methods and may be: - ``_start_TAG`` called when the start tag is found - ``_end_TAG`` called when the end tag is found rcCst||tj|_|j||jtjjj d|jtjjj d|jtjjj d|jtjjj dd|_ |dS)zuInitialize the parser. Arguments: - debug - integer, amount of debug information to print rrN)rrxmlZsaxZ make_parserZ_parserZsetContentHandlerZ setFeatureZhandlerZfeature_validationZfeature_namespacesZfeature_external_pesZfeature_external_ges _xml_versionreset)r rrrrrs   zBlastParser.__init__cCs&g|_t|_t|_d|j_dS)z>Reset all the data allowing reuse of the BlastParser() object.N)_recordsrZHeader_headerZ Parameters _parametersfilter)r rrrr,s  zBlastParser.resetcCs4|dkr|n|dkr$|n td|dS)NZ BlastOutputZ BlastXML2zOInvalid root node name: %s. Root node should be either BlastOutput or BlastXML2)_setup_blast_v1_setup_blast_v2r)r rrrrrs  zBlastParser._on_root_nodec3Cs|j|j|j|j|j|j|j|j|j|j |j |j |j |j |j|j|j|j|j|j|j|j|j|j|j|j|j|j|j|j|j|j|j |j!|j"|j#|j$|j%|j&|j'|j(|j)|j*|j+|j,|j-|j.|j/|j0|j1d2|_2dS)N)2Zstart_IterationZ end_IterationZend_BlastOutput_programZend_BlastOutput_versionZend_BlastOutput_referenceZend_BlastOutput_dbzend_BlastOutput_query-IDzend_BlastOutput_query-defzend_BlastOutput_query-lenzend_Iteration_query-IDzend_Iteration_query-defzend_Iteration_query-lenend_BlastOutput_hitsZend_Parameters_matrixZend_Parameters_expectzend_Parameters_sc-matchzend_Parameters_sc-mismatchzend_Parameters_gap-openzend_Parameters_gap-extendZend_Parameters_filterZ start_HitZend_HitZ end_Hit_idZ end_Hit_defZend_Hit_accessionZ end_Hit_lenZ start_HspZ end_Hsp_scorezend_Hsp_bit-scoreZend_Hsp_evaluezend_Hsp_query-fromzend_Hsp_query-tozend_Hsp_hit-fromzend_Hsp_hit-tozend_Hsp_query-framezend_Hsp_hit-frameZend_Hsp_identityZend_Hsp_positiveZ end_Hsp_gapszend_Hsp_align-lenZ end_Hsp_qseqZ end_Hsp_hseqZend_Hsp_midlinezend_Statistics_db-numzend_Statistics_db-lenzend_Statistics_hsp-lenzend_Statistics_eff-spaceZend_Statistics_kappaZend_Statistics_lambdaZend_Statistics_entropy)3_start_blast_record_end_blast_record_set_header_application_set_header_version_set_header_reference_set_header_database_set_header_query_id_set_header_query_set_header_query_letters_set_record_query_id_set_record_query_def_set_record_query_letters_set_record_hits_set_parameters_matrix_set_parameters_expect_set_parameters_sc_match_set_parameters_sc_mismatch_set_parameters_gap_penalties_set_parameters_gap_extend_set_parameters_filter _start_hit_end_hit set_hit_id set_hit_defset_hit_accession set_hit_len _start_hsp_set_hsp_score_set_hsp_bit_score_set_hsp_e_value_set_hsp_query_start_set_hsp_query_end_set_hsp_hit_from_set_hsp_hit_to_set_hsp_query_frame_set_hsp_hit_frame_set_hsp_identity_set_hsp_positive _set_hsp_gaps_set_hsp_align_len_set_hsp_query_seq_set_hsp_subject_seq_set_hsp_midline_set_statistics_db_num_set_statistics_db_len_set_statistics_hsp_len_set_statistics_eff_space_set_statistics_kappa_set_statistics_lambda_set_statistics_entropyr )r rrrr1sdzBlastParser._setup_blast_v1c7Csd|_d|_|j|j|j|j|j|j|j|j |j |j |j |j |j|j|j|j|j|j|j|j|j|j|j|j|j|j|j|j|j|j|j |j!|j"|j#|j$|j%|j&|j'|j(|j)|j*|j+|j,|j-|j.|j/|j0|j1|j2|j3|j4|j5|j6|j7d6|_8dS)Nr )6zstart_report/Reportzend_report/Reportzend_Report/programzend_Report/versionzend_Report/referencez end_Target/dbzend_Search/query-idzend_Search/query-titlezend_Search/query-lenr3zend_Parameters/matrixzend_Parameters/expectzend_Parameters/sc-matchzend_Parameters/sc-mismatchzend_Parameters/gap-openzend_Parameters/gap-extendzend_Parameters/filterzstart_hits/Hitz end_hits/Hitzstart_description/HitDescrzend_description/HitDescrzend_HitDescr/idzend_HitDescr/accessionzend_HitDescr/titlezend_HitDescr/taxidzend_HitDescr/scinamez end_Hit/lenzstart_hsps/Hspz end_hsps/Hspz end_Hsp/scorezend_Hsp/bit-scorezend_Hsp/evaluezend_Hsp/query-fromzend_Hsp/query-tozend_Hsp/hit-fromzend_Hsp/hit-tozend_Hsp/query-framezend_Hsp/hit-framezend_Hsp/query-strandzend_Hsp/hit-strandzend_Hsp/identityzend_Hsp/positivez end_Hsp/gapszend_Hsp/align-lenz end_Hsp/qseqz end_Hsp/hseqzend_Hsp/midlinezend_Statistics/db-numzend_Statistics/db-lenzend_Statistics/hsp-lenzend_Statistics/eff-spacezend_Statistics/kappazend_Statistics/lambdazend_Statistics/entropy)9r r+r4r5r6r7r8r9r=r>r?r@rArBrCrDrErFrGrHrI_start_hit_descr_item_end_hit_descr_item_end_description_id_end_description_accession_end_description_title_end_description_taxid_end_description_scinamerMrN_end_hsprOrPrQrRrSrTrUrVrW_set_hsp_query_strand_set_hsp_hit_strandrXrYrZr[r\r]r^r_r`rarbrcrdrer )r rrrr2spzBlastParser._setup_blast_v2cCst|_dS)zStart interaction (PRIVATE).N)rZBlast_blast)r rrrr4(szBlastParser._start_blast_recordcCs<|jj|j_|jj|j_|jj|j_|jj|j_|jj|j_t|jdrP|jjs\|jj|j_t|jdrp|jj s||jj |j_ t|jdr|jj s|jj |j_ |jj |j_ |jj |j_ |jj|j_|jj|j_|jj|j_|jj|j_|jj|j_|jj|j_|jj|j_|jj|j_|j|jd|_|jr8tddS)zEnd interaction (PRIVATE).queryquery_id query_lettersNz&NCBIXML: Added Blast record to results)r. referencerpdateversiondatabase applicationhasattrrqrrrsZ query_lengthnum_letters_in_databaseZdatabase_lengthnum_sequences_in_databaseZdatabase_sequencesr/matrixZnum_seqs_better_e gap_penaltiesr0expectsc_match sc_mismatchr-rr r)r rrrr5,s2                  zBlastParser._end_blast_recordcCs|j|j_dS)zsBLAST program, e.g., blastp, blastn, etc. (PRIVATE). Save this to put on each blast record object N)rupperr.rx)r rrrr6dsz#BlastParser._set_header_applicationcCsh|j}|d|j_t|dkrd|dddkrX|dddkrX|ddd|j_n |d|j_dS) zVersion number and date of the BLAST engine (PRIVATE). e.g. "BLASTX 2.2.12 [Aug-07-2005]" but there can also be variants like "BLASTP 2.2.18+" without the date. Save this to put on each blast record object rrr r[]N)rsplitr.rvrru)r partsrrrr7ks     zBlastParser._set_header_versioncCs|j|j_dS)zRecord any article reference describing the algorithm (PRIVATE). Save this to put on each blast record object N)rr.rt)r rrrr8sz!BlastParser._set_header_referencecCs|j|j_dS)ziRecord the database(s) searched (PRIVATE). Save this to put on each blast record object N)rr.rw)r rrrr9sz BlastParser._set_header_databasecCs|j|j_dS)zRecord the identifier of the query (PRIVATE). Important in old pre 2.2.14 BLAST, for recent versions is enough N)rr.rr)r rrrr:sz BlastParser._set_header_query_idcCs|j|j_dS)zRecord the definition line of the query (PRIVATE). Important in old pre 2.2.14 BLAST, for recent versions is enough N)rr.rq)r rrrr;szBlastParser._set_header_querycCst|j|j_dS)zRecord the length of the query (PRIVATE). Important in old pre 2.2.14 BLAST, for recent versions is enough N)intrr.rs)r rrrr<sz%BlastParser._set_header_query_letterscCs|j|j_dS)z-Record the identifier of the query (PRIVATE).N)rrprr)r rrrr=sz BlastParser._set_record_query_idcCs|j|j_dS)z2Record the definition line of the query (PRIVATE).N)rrprq)r rrrr>sz!BlastParser._set_record_query_defcCst|j|j_dS)z)Record the length of the query (PRIVATE).N)rrrprs)r rrrr?sz%BlastParser._set_record_query_letterscCst|j|j_dS)zAHits to the database sequences, one for every sequence (PRIVATE).N)rrrpZnum_hits)r rrrr@szBlastParser._set_record_hitscCs|j|j_dS)z+Matrix used (-M on legacy BLAST) (PRIVATE).N)rr/r|)r rrrrAsz"BlastParser._set_parameters_matrixcCs|j|j_dS)zExpect values cutoff (PRIVATE).N)rr/r~)r rrrrBs z"BlastParser._set_parameters_expectcCst|j|j_dS)z@Match score for nucleotide-nucleotide comparison (-r) (PRIVATE).N)rrr/r)r rrrrCsz$BlastParser._set_parameters_sc_matchcCst|j|j_dS)zEMismatch penalty for nucleotide-nucleotide comparison (-r) (PRIVATE).N)rrr/r)r rrrrDsz'BlastParser._set_parameters_sc_mismatchcCst|j|j_dS)z"Gap existence cost (-G) (PRIVATE).N)rrr/r})r rrrrEsz)BlastParser._set_parameters_gap_penaltiescCs|jjt|jf|j_dS)z"Gap extension cose (-E) (PRIVATE).N)r/r}rr)r rrrrFsz&BlastParser._set_parameters_gap_extendcCs|j|j_dS)z(Record filtering options (-F) (PRIVATE).N)rr/r0)r rrrrGsz"BlastParser._set_parameters_filtercCs`|jjt|jdkr$tnt|_|jj |jg|j_ |jjd|_ d|j_ dS)z Start filling records (PRIVATE).rrrN) rpZ alignmentsrrZ Alignmentr+Z DescriptionZDescriptionExt_descrZ descriptionsmultiple_alignment_hitnum_alignments)r rrrrHs zBlastParser._start_hitcCsd|j_d|_d|_dS)zClear variables (PRIVATE).N)rprrr)r rrrrIszBlastParser._end_hitcCs|j|j_|jd|j_dS)z9Record the identifier of the database sequence (PRIVATE). N)rrhit_idtitle)r rrrrJ s zBlastParser.set_hit_idcCs,|j|j_|jj|j7_|jj|j_dS)z>Record the definition line of the database sequence (PRIVATE).N)rrZhit_defrr)r rrrrKs zBlastParser.set_hit_defcCs|j|j_|j|j_dS)z>Record the accession value of the database sequence (PRIVATE).N)rr accessionr)r rrrrLs zBlastParser.set_hit_accessioncCst|j|j_dS)zRecord the length of the hit.N)rrrlength)r rrrrMszBlastParser.set_hit_lencCsVt|_d|j_|jj|j|jjd7_|j j t |j j d|_ dS)Nrr) rZHSP_hsp positivesrZhspsrrrrprZMultipleAlignmentZ_mult_al)r rrrrN"s  zBlastParser._start_hspcCs,|jjr(t|jjdkr(|jjd7_dS)Nr)r)rframer)r rrrrm,szBlastParser._end_hspcCs,t|j|j_|jjdkr(t|j|j_dS)z&Record the raw score of HSP (PRIVATE).N)floatrrZscorer)r rrrrO1s zBlastParser._set_hsp_scorecCs,t|j|j_|jjdkr(t|j|j_dS)z&Record the Bit score of HSP (PRIVATE).N)rrrbitsr)r rrrrP7s zBlastParser._set_hsp_bit_scorecCs,t|j|j_|jjdkr(t|j|j_dS)z-Record the expect value of the HSP (PRIVATE).N)rrrr~re)r rrrrQ=s zBlastParser._set_hsp_e_valuecCst|j|j_dS)zEOffset of query at the start of the alignment (one-offset) (PRIVATE).N)rrrZ query_start)r rrrrRCsz BlastParser._set_hsp_query_startcCst|j|j_dS)zCOffset of query at the end of the alignment (one-offset) (PRIVATE).N)rrrZ query_end)r rrrrSGszBlastParser._set_hsp_query_endcCst|j|j_dS)zLOffset of the database at the start of the alignment (one-offset) (PRIVATE).N)rrrZ sbjct_start)r rrrrTKszBlastParser._set_hsp_hit_fromcCst|j|j_dS)zJOffset of the database at the end of the alignment (one-offset) (PRIVATE).N)rrrZ sbjct_end)r rrrrUOszBlastParser._set_hsp_hit_tocCs:t|j}|f|j_|jjdkr6|dkr,dndf|j_dS)z+Frame of the query if applicable (PRIVATE).BLASTNrPlusMinusN)rrrrr.rxstrand)r vrrrrV[s   z BlastParser._set_hsp_query_framecCsht|j}t|jjdkr(d|f|j_n|jj|f7_|jjdkrd|jj|dkrZdndf7_dS)z7Frame of the database sequence if applicable (PRIVATE).rrrrN)rrrrrr.rxr)r rrrrrWbs   zBlastParser._set_hsp_hit_framecCs4|jf|j_|jjdkr0|jdkr&dndf|j_dS)z+Frame of the query if applicable (PRIVATE).rrrrN)rrrr.rxr)r rrrrnls  z!BlastParser._set_hsp_query_strandcCsD|jj|jf7_|jjdkr@|jj|jdkr6dndf7_dS)z7Frame of the database sequence if applicable (PRIVATE).rrrrN)rrrr.rxr)r rrrrors zBlastParser._set_hsp_hit_strandcCs*t|j}||j_|jjdkr&||j_dS)z;Record the number of identities in the alignment (PRIVATE).N)rrrZ identitiesr)r rrrrrXxs  zBlastParser._set_hsp_identitycCst|j|j_dS)zVRecord the number of positive (conservative) substitutions in the alignment (PRIVATE).N)rrrr)r rrrrYszBlastParser._set_hsp_positivecCst|j|j_dS)z5Record the number of gaps in the alignment (PRIVATE).N)rrrZgaps)r rrrrZszBlastParser._set_hsp_gapscCst|j|j_dS)z-Record the length of the alignment (PRIVATE).N)rrrZ align_length)r rrrr[szBlastParser._set_hsp_align_lencCs|j|j_dS)z4Record the alignment string for the query (PRIVATE).N)rrrq)r rrrr\szBlastParser._set_hsp_query_seqcCs|j|j_dS)z7Record the alignment string for the database (PRIVATE).N)rrsbjct)r rrrr]sz BlastParser._set_hsp_subject_seqcCsF|j|j_t|jjt|jjks&tt|jjt|jjksBtdS)zBRecord the middle line as normally seen in BLAST report (PRIVATE).N)rrmatchrrqAssertionErrorr)r rrrr^s zBlastParser._set_hsp_midlinecCst|j|j_dS)z9Record the number of sequences in the database (PRIVATE).N)rrrpr{)r rrrr_sz"BlastParser._set_statistics_db_numcCst|j|j_dS)z7Record the number of letters in the database (PRIVATE).N)rrrprz)r rrrr`sz"BlastParser._set_statistics_db_lencCst|j|j_dS)z*Record the effective HSP length (PRIVATE).N)rrrpZeffective_hsp_length)r rrrrasz#BlastParser._set_statistics_hsp_lencCst|j|j_dS)z,Record the effective search space (PRIVATE).N)rrrpZeffective_search_space)r rrrrbsz%BlastParser._set_statistics_eff_spacecCst|j|j_dS)z&Karlin-Altschul parameter K (PRIVATE).N)rrrp ka_params)r rrrrcsz!BlastParser._set_statistics_kappacCst|j|jjf|j_dS)z+Karlin-Altschul parameter Lambda (PRIVATE).N)rrrpr)r rrrrdsz"BlastParser._set_statistics_lambdacCs|jjt|jf|j_dS)z&Karlin-Altschul parameter H (PRIVATE).N)rprrr)r rrrresz#BlastParser._set_statistics_entropycCst|_dS)z#XML v2. Start hit description item.N)rZDescriptionExtItem_hit_descr_item)r rrrrfsz!BlastParser._start_hit_descr_itemcCs.|j|j|jjs$t|j|j_d|_dS)z#XML v2. Start hit description item.N)rZ append_itemrrrstr)r rrrrgszBlastParser._end_hit_descr_itemcCs |j|j_|jjs|j|j_dS)z9XML v2. The identifier of the database sequence(PRIVATE).N)rridrr)r rrrrhs zBlastParser._end_description_idcCs&|j|j_t|jdds"|j|j_dS)z?XML v2. The accession value of the database sequence (PRIVATE).rN)rrrgetattrr)r rrrris z&BlastParser._end_description_accessioncCs|j|j_dS)z,XML v2. The hit description title (PRIVATE).N)rrr)r rrrrjsz"BlastParser._end_description_titlecCs,yt|j|j_Wntk r&YnXdS)N)rrrZtaxidr)r rrrrksz"BlastParser._end_description_taxidcCs|j|j_dS)N)rrZsciname)r rrrrlsz$BlastParser._end_description_scinameN)r)Er%r&r'r(rr,rr1r2r4r5r6r7r8r9r:r;r<r=r>r?r@rArBrCrDrErFrGrHrIrJrKrLrMrNrmrOrPrQrRrSrTrUrVrWrnrorXrYrZr[r\r]r^r_r`rarbrcrdrerfrgrhrirjrkrlrrrrr)~s   6<8      r)cCsdt||}y t|}Wntk r4tddYnXyt|tdWntk r^YnX|S)aOReturn a single Blast record (assumes just one query). Uses the BlastParser internally. This function is for use when there is one and only one BLAST result in your XML file. Use the Bio.Blast.NCBIXML.parse() function if you expect more than one BLAST record (i.e. if you have more than one query sequence). zNo records found in handleNz$More than one record found in handle)parsenext StopIterationr)handleriteratorrecordrrrreads   rc csddlm}d}d}d}d}d}d}||} t| trHd}d }d }d }| sTtd x| r| |std || d df|} t|} | j | _ | j | _ | j | _| | dx*| jr| jd} | jdd | _| VqWx|||d} }| s | |dP||}||| |krR| | dxt| jrN| jdVq4Wq| |||d\} }||}| | dx| jr| jdVqW||} }PqW|rt|t| jdksXtt| jqXW| rt| |rt|t| jdkstt| jd S)aReturn an iterator a Blast record for each query. Incremental parser, this is an iterator that returns Blast records. It uses the BlastParser internally. handle - file handle to and XML file to parse debug - integer, amount of debug information to print This is a generator function that returns multiple Blast records objects - one for each query sequence given to blast. The file is read incrementally, returning complete records as they are read in. Should cope with new BLAST 2.2.14+ which gives a single XML file for multiple query records. Should also cope with XML output from older versions BLAST which gave multiple XML files concatenated together (giving a single file which strictly speaking wasn't valid XML). r)expati z s  je