ó ª ^Ic)@sdZddlZyddlZWnek rAddlZnXddlTdefd„ƒYZdfd„ƒYZe e e e dd e e e d e e e e e e e e e e e e e e e e e e d e d e e e e e d e e e d „(Z d„Z dS(s4 This module provides code to work with the WWW version of BLAST provided by the NCBI. http://blast.ncbi.nlm.nih.gov/ Functions: qblast Do a BLAST search using the QBLAST API. Deprecated classes: BlastParser Parses output from WWW blast. _Scanner Scans output from NCBI's BLAST WWW server. iÿÿÿÿN(t*t BlastParsercBs eZdZd„Zd„ZRS(s†Parses WWW BLAST data into a Record.Blast object (DEPRECATED). This is a parser for the NCBI's HTML (web page) BLAST output. cCsUddl}|jdddtƒddl}tƒ|_t|jƒƒ|_dS(s)Create a BlastParser object (DEPRECATED).iÿÿÿÿNs,Bio.Blast.NCBIWWW.BlastParser is deprecated.s) We recommend you use the XML output withs) the parser in Bio.Blast.NCBIXML instead.( twarningstwarntDeprecationWarningtNCBIStandalonet_Scannert_scannertSGMLStrippingConsumert_BlastConsumert _consumer(tselfRR((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyt__init__%s    cCs |jj||jƒ|jjS(sparse(self, handle)(RtfeedR tdata(R thandle((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pytparse1s(t__name__t __module__t__doc__R R(((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR s RcBs°eZdZd„Zd„Zd„Zd„Zd„Zd„Zd„Z d„Z d „Z d „Z d „Z d „Zd „Zd„Zd„Zd„Zd„Zd„ZRS(sôScanner for the HTML BLAST parser (PRIVATE, DEPRECATED). Scan BLAST output from NCBI's web server at: http://www.ncbi.nlm.nih.gov/BLAST/ Tested with BLAST v2.0.10 Methods: feed Feed data into the scanner. cCsÀddlm}t||jƒr+|}n|j|ƒ}t||jdtjdƒƒ|j||ƒ|j ||ƒ|j ||ƒ|j ||ƒx |j ƒr»t ||jƒqœWdS(sðS.feed(handle, consumer) Feed in a BLAST report for scanning. handle is a file-like object that contains the BLAST report. consumer is a Consumer object that will receive events as the report is scanned. iÿÿÿÿ(tFilethas_res .?BLASTN(tBioRt isinstancet UndoHandletread_and_call_untiltnoeventtretcompilet _scan_headert _scan_roundst_scan_database_reportt_scan_parameterstpeeklinet read_and_call(R RtconsumerRtuhandle((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR As cCs,|jƒt||jddƒt||jddƒxK|jƒ}|d dksb|jƒ rs|j|ƒPn|j|ƒq9Wt||jddƒt||jƒ|j ƒj d ƒd krè|j ||ƒ|j ||ƒn |j ||ƒ|j ||ƒt||jddƒ|j ƒdS( NtcontainstBLASTtblankiis

tstarttRIDsQuery=i(t start_headerR"tversiontread_and_call_whileRtreadlinetstript referencetattempt_read_and_callR!tfindt_scan_query_infot_scan_database_infot end_header(R R$R#tline((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRrs$"   cCset||jddƒraxF|jƒ}t|ƒr7Pnd|krGPn|j|jƒƒqWndS(NR%t BLASTFORMsQuery=(R0RR!t is_blank_lineR-(R R$R#R5((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyt_scan_blastform¶s    cCsÿt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒ|j||ƒt||jddƒt||jdd ƒrt||jdd ƒ|j|jƒƒt||jddƒnt||jddƒt||jdd ƒrKt||jdd ƒnt||jdd ƒt||jddƒråt||jdd ƒxS|jƒ}|d d k sÄ|jdƒdkrÈPnt||jdd ƒq’Wnt ||jddƒdS(NR(s

R%tDatabases sequences;R'isproblems or questionsssTaxonomy reportss
s
s
isQuery=i( R0RR"t database_infoRR8R-R!R1R,(R R$R#R5((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR3Ás:    &cCs‡t||jddƒt||jddƒt||jddƒt||jddƒrst||jddƒn|j||ƒdS(NR%sQuery=R'iR(s
(R"t
query_infoRR,RR0R8(RR$R#((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR2ïscCs$|j||ƒ|j||ƒdS(N(t_scan_descriptionst_scan_alignments(RR$R#((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRøscCsì|jƒt||jdtjdƒƒset||jddƒt||jddƒ|jƒdSt	||jddƒt	||jddƒt||j
dd	dd
ƒt||jddƒsÞt||jddƒn|jƒdS(NRsScore {4,5}ER%sNo significant similarityR'iR(sSequences producingis(tstart_descriptionsR0tdescription_headerRRtno_hitsR,Rtend_descriptionsR"tdescription(RR$R#((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR<üs$


	cCsñt|ƒ}t|ƒ}|j|ƒ|j|ƒd}}d|krQd}nj|jdƒrcnX|jdƒrunF|jdƒrd}n.|jdƒr¥d}ntd||fƒ‚|rÔ|j||ƒn|rí|j||ƒndS(	Nit
Alignmentsis
  DatabasesLambda     K      Ht	blast_tmps
s'Cannot resolve location at lines:
%s
%s(t
safe_readlinetsavelinet
startswitht
ValueErrort_scan_pairwise_alignmentst_scan_masterslave_alignment(RR$R#tline1tline2tis_pairwisetis_masterslave((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR=$s*


			cCsÉxÂt||jddƒt|ƒ}t|ƒ}|j|ƒ|j|ƒ|jdƒdksŠ|jdƒdksŠ|jdƒdkrŽPn|d dkr±|j||ƒq|j||ƒqWdS(NR(s
R9isLambda     K      HisQuery:(RRRERFR1t$_scan_abbreviated_pairwise_alignmentt_scan_one_pairwise_alignment(RR$R#RKRL((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRIWs

*cCsR|jƒ|jƒt||jddƒ|j||ƒ|jƒ|jƒdS(NR(s
(tstart_alignmentt	start_hspR"Rt_scan_hsp_alignmenttend_hspt
end_alignment(RR$R#((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyROns	


cCs|jƒt||jddƒ|j||ƒxØt|ƒ}t|ƒ}t|ƒ}|j|ƒ|j|ƒ|j|ƒtjdƒ}|jd|ƒ}|jd|ƒ}|jd|ƒ}|d dkr÷|d dkr÷|d dkr÷Pn|j	||ƒq3W|j
ƒdS(NR(s
s
]*>tis Score(RQR"Rt_scan_alignment_headerRERFRRtsubt	_scan_hspRU(RR$R#RKRLtline3trea((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRP~s&



 cCs–x^t|ƒ}|jƒjdƒr5|j|ƒPnt|ƒrPtdƒ‚n|j|ƒqWt||jddƒs’t	||jddƒndS(NsLength =s*I missed the Length in an alignment headerR(s
          R'i(
REtlstripRGtlengthR7RHttitleR0RR"(RR$R#R5((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRW¦s
	cCs8|jƒ|j||ƒ|j||ƒ|jƒdS(N(RRt_scan_hsp_headerRSRT(RR$R#((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRY¸s
cCs§t||jddƒt||jddƒt||jdtjdƒƒt||jddƒt||jddƒt||jdd	ƒt||jddƒdS(
NR(s
R'iRs^ (]*>)*Scores Identitiess Strands Frame(	R0RR"tscoreRRt
identitieststrandtframe(RR$R#((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR_¾s
cCs«xxt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒsPqqWt||jddƒt||jddƒdS(NR(s     tQuerytSbjctR'is
(R0RR"tquerytaligntsbjctR,(RR$R#((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRSÓscCs¡|jƒt||jddƒxZt|ƒ}t|ƒrK|j|ƒq#|d dkrl|j|ƒPq#|j|ƒq#Wt||jddƒ|jƒdS(NR(s
is
R'i(RQR"RRER7t	multalignR,RU(RR$R#R5((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRJês

cCs|jƒ|jƒ}|jdƒdkrDt||jddƒn|jƒ}|jdƒdkrìt||jddƒt||jddƒt||jdd	ƒt||jdd
ƒt||j	ddƒt||jddƒnt||jdd
ƒt||j
ƒt||jddƒt||jddƒt||jdd
ƒrit||j
ƒnt||jddƒ|jƒdS(NR9iR(s
sLambda     K      HR%s
  DatabasetPosteds
    Posteds  Number of letterss  Number of sequencess  tLambdaR'itGapped(tstart_database_reportR!R1R"RtdatabaseRtposted_datetnum_letters_in_databasetnum_sequences_in_databaset	ka_paramsR0tgappedt
ka_params_gapR,tend_database_report(RR$R#R5RL((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRùs,
cCs²|jƒt||jddƒt||jddƒt||jddƒrgt||jddƒn,t||jddƒt||jddƒt||jddƒt||jddƒt||j	ddƒt||j
dd	ƒryt||jdd
ƒr4t||jddƒt||j
ddƒqyt||jddƒt||jdd
ƒt||jdd
ƒnt||jddƒrt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒnšt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jddƒt||jdd ƒt||jdd!ƒt||jdd"ƒt||jdd#ƒt||jd$d%ƒt||jdd&ƒt||jdd'ƒ|jƒdS((NR(tMatrixtGapsNumber of HitssNumber of SequencessNumber of extensionssNumber of successfulsNumber of sequencessNumber of HSP's bettersNumber of HSP's successfullysNumber of HSP's thatsNumber of HSP's gappedsNumber of extra gappedsLength of querysLength of databasesLength adjustmentsEffective length of querysEffective length of databasesEffective search space:sEffective search space usedslength of queryslength of databases
effective HSPseffective length of queryseffective length of databaseseffective search space:seffective search space usedt
frameshifttTtAtX1tX2tX3tS1tS2R'is
s( tstart_parametersR0tmatrixt
gap_penaltiestnum_hitsR"t
num_sequencestnum_extendstnum_good_extendstnum_seqs_better_ethsps_no_gapthsps_prelim_gappedthsps_prelim_gap_attemptedthsps_gappedtno_eventtquery_lengthtdatabase_lengthteffective_query_lengthteffective_database_lengthteffective_search_spaceteffective_search_space_usedteffective_hsp_lengthRxt	thresholdtwindow_sizetdropoff_1st_passt
gap_x_dropofftgap_x_dropoff_finaltgap_triggertblast_cutoffRtend_parameters(RR$R#((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR ,s
	
		

	
(RRRR
RR8R3R2RR<R=RIRORPRWRYR_RSRJRR (((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyR6s&
	1	D		.				(	3			(						3s(none)g$@i2iôtXMLc+>!Cs ddl}+ddl},ddl}-|d@ks6t‚d|fd	|fd
|fd|fd|fd
|fd|fd|	fd|
fd|fd|fd|
fd|fd|fd|fd|fd|fd|fd|fd|fd|fd|fd|fd|fd |fd!|fd"|fd#|fd$|fd%|fd&|fdAg }.g|.D]}/|/d)dk	r]|/^q]}0|+j|0ƒ}1|,jd*|1id+d,6ƒ}2|,j|2ƒ}3t|3ƒ\}4}5d-|fd.| fd/|!fd0|"fd1|#fd2|$fd3|%fd4|&fd5|'fd6|(fd7|4fd8|)fd#|fd9|*fdBg}.g|.D]}/|/d)dk	rY|/^qY}0|+j|0ƒ}1d;}6|-jƒ}7xôt	r’|-jƒ}8|7|6|8}9|9d<krå|-j
|9ƒ|8|9}7n|8}7|,jd*|1id+d,6ƒ}2|,j|2ƒ}3|3jƒ}:|:jd=ƒd<kr;Pn|:j
d=ƒ};|:j
d>|;ƒ}<|:|;td=ƒ|<!jƒ}=|=jƒd?krŸPqŸqŸWtj|:ƒS(CsHDo a BLAST search using the QBLAST server at NCBI.

    Supports all parameters of the qblast API for Put and Get.
    Some useful parameters:
    program        blastn, blastp, blastx, tblastn, or tblastx (lower case)
    database       Which database to search against (e.g. "nr").
    sequence       The sequence to search.
    ncbi_gi        TRUE/FALSE whether to give 'gi' identifier.
    descriptions   Number of descriptions to show.  Def 500.
    alignments     Number of alignments to show.  Def 500.
    expect         An expect value cutoff.  Def 10.0.
    matrix_name    Specify an alt. matrix (PAM30, PAM70, BLOSUM80, BLOSUM45).
    filter         "none" turns off filtering.  Default no filtering
    format_type    "HTML", "Text", "ASN.1", or "XML".  Def. "XML".
    entrez_query   Entrez query to limit Blast search
    hitlist_size   Number of hits to return. Default 50

    This function does no checking of the validity of the parameters
    and passes the values to the server as is.  More help is available at:
    http://www.ncbi.nlm.nih.gov/BLAST/blast_overview.html

    iÿÿÿÿNtblastntblastptblastxttblastnttblastxtAUTO_FORMATtCOMPOSITION_BASED_STATISTICStDATABASEtDB_GENETIC_CODEt	ENDPOINTStENTREZ_QUERYtEXPECTtFILTERtGAPCOSTStGENETIC_CODEtHITLIST_SIZEtI_THRESHtLAYOUTt
LCASE_MASKtMATRIX_NAMEtNUCL_PENALTYtNUCL_REWARDtOTHER_ADVANCEDt
PERC_IDENTtPHI_PATTERNtPROGRAMtQUERYt
QUERY_FILEtQUERY_BELIEVE_DEFLINEt
QUERY_FROMtQUERY_TOtSEARCHSP_EFFtSERVICEt	THRESHOLDtUNGAPPED_ALIGNMENTt	WORD_SIZEtCMDtPutis'http://blast.ncbi.nlm.nih.gov/Blast.cgitBiopythonClients
User-Agentt
ALIGNMENTStALIGNMENT_VIEWtDESCRIPTIONStENTREZ_LINKS_NEW_WINDOWt
EXPECT_LOWtEXPECT_HIGHtFORMAT_ENTREZ_QUERYt
FORMAT_OBJECTtFORMAT_TYPEtNCBI_GIR)tRESULTS_FILEt
SHOW_OVERVIEWtGetg@isStatus=s
tREADY(RRžRŸR R¡(RÁRÂ(RÁRÐ(turllibturllib2ttimetAssertionErrortNonet	urlencodetRequestturlopent_parse_qblast_ref_pagetTruetsleeptreadR1tindextlenR.tuppertStringIO(>tprogramRntsequencetauto_formattcomposition_based_statisticstdb_genetic_codet	endpointstentrez_querytexpecttfiltertgapcoststgenetic_codethitlist_sizeti_threshtlayoutt
lcase_masktmatrix_nametnucl_penaltytnucl_rewardtother_advancedt
perc_identtphi_patternt
query_filetquery_believe_deflinet
query_fromtquery_totsearchsp_efftserviceR”tungapped_alignmentt	word_sizet
alignmentstalignment_viewtdescriptionstentrez_links_new_windowt
expect_lowtexpect_hightformat_entrez_queryt
format_objecttformat_typetncbi_gitresults_filet
show_overviewRÒRÓRÔt
parameterstxRftmessagetrequestRtridtrtoetdelaytprevioustcurrenttwaittresultstitjtstatus((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pytqblast¦s¢%																																)																)	

	cCsS|jƒ}|jdƒ}|dkr0d
}n/|jd|ƒ}||tdƒ|!jƒ}|jdƒ}|dkrƒd
}n/|jd|ƒ}||tdƒ|!jƒ}|rÏ|rÏtdƒ‚n>|sîtdt|ƒƒ‚n|s
tdt|ƒƒ‚ny|t|ƒfSWn+tk
rNtdd	t|ƒƒ‚nXd
S(sÛExtract a tuple of RID, RTOE from the 'please wait' page (PRIVATE).

    The NCBI FAQ pages use TOE for 'Time of Execution', so RTOE is proably
    'Request Time of Execution' and RID would be 'Request Identifier'.
    sRID =iÿÿÿÿs
sRTOE =sdNo RID and no RTOE found in the 'please wait' page. (there was probably a problem with your request)s<No RID found in the 'please wait' page. (although RTOE = %s)s<No RTOE found in the 'please wait' page. (although RID = %s)sA non-integer RTOE found in sthe 'please wait' page, %sN(RÝR1RÖRßR.RHtreprtint(RtsRRRR((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyRÚ3s0		
(RRt	cStringIORátImportErrortBio.ParserSupporttAbstractParserRRRÖRRÚ(((s„/oak/stanford/groups/akundaje/marinovg/programs/biopython-1.50.tar.gz/biopython-1.50/build/lib.linux-x86_64-2.7/Bio/Blast/NCBIWWW.pyts0

ÿÿs							€