1、會(huì)計(jì)學(xué)1大規(guī)?；蚪M測(cè)序中的信息分析大規(guī)模基因組測(cè)序中的信息分析Sequence Data DistributionVectorSequencing and Data Processing Procedure BasecallingRepeat MarkORF Prediction Gene AnnotationFinishingAssembleVector MarkPhredPhd2fastaCrossmatchPhrapConsedRepeatmaskerGlimmerBlastxBlastnClastaltRNAscanQualityControlQualCalQualDrawQualS

2、tatCOGsSwiss-port,PIR, GDB,GenBankSequencing-xl-1yW =synaptic weight vector of a neuron in layer l =threshold of a neuron in layer lV =vector of net internal activity levels of neurons Y =vector of function signals of neurons in layer l (l)(l)(l)(l)vjlnwjilipnyiln()()()()()01()yvjljlnn()()( )( )exp(

3、)11 Forward Computationinput vector xdesired response vextor d The net internal activity level for neuron j in layer l isIf neuron j is in the output layer (i.e.,l=L), set yojLjnn()()()Hence, compute the error signaledojjj Backward Computationvector of local gradients of neurons in layer le = error

4、vector respresented by e , e , ., e as elements12qjljljljlkjlnynynnn()()()()()()()()()()1jLjLjjnen onon()()()()()()1for neuron j in output layer Lfor neuron j in hidden layer l()1wnwnwnwnn ynjiljiljiljiljljl()()()()()()()( )( )()( )( )111 Iteration-GU -AG-splicing sites exonexonintronpseudosplicing

5、sites-GU -AG -Numbe of splicing sites andpseudosplicing sites of testing groupsplicing sitespseudosplicingGU601710AG602800Distribution Matrix of Prediction38226015278528005328072860success rate = 38/60 64%DNNNRRNNNNNNN000111221 20()ln/ ln,()/RNNRNNNNi iNiNN2020000111,()D is the fractal dimension and

6、 there exists self-similarity in this rangeof the nucleotide sequence.CGAT-CGCGGCGTGTGTTATA - 0 0.5 1 1.5 2 2.5 3 3.5 07654321Intron Seq.Exon Seq.Random Seq.End-to-end Rangel n/RN212Main RangeGeometric RangelnNFractal dimension of classes ofsequencessequenceDgDgDmDmEX1.9040.1031.6640.230RE1.9530.072

7、1.7730.274IN1.7110.1471.5910.216RI1.9350.1341.6680.309 and are average of and of each class of sequences, and are their standard deviations. The class of exon sequences is denoted as EX, and intron sequences IN, randomized exon sequences RE, and randomized intron sequences RI.gDDmgDDmDgDmComplexity

8、analysisUsing the methods and techniques in cryptology Coincident indexICCCffNNfNlLlllLl220011()()The original presentation of coincident index is:in which N is the number of characters in the sequence,L is the number of different kinds of character, and is the frequency of the lth character appeari

9、ng in this sequence.flThe coincident index introduced from cryptology is applied to DNA sequence analysis. In case of mono-letter,the expressing of coincident index IC of DNA sequence is:ICSTPILbbRIPjkjkkj()()1641j=0,1,.,L-1; k=A,C,G,T.R=4 25 (25 1)=2400 In case of biletter,the expression of IC is:I

10、CSTPILbbRIPjkjkkj()()1641j=0,1,.,L-1; k=AA,AC,AG,AT,CA,CC,.TG,TT R1 61 0 01 61 0 01 615 2 5()Composite Artificial Neural Network SystemAuto-annotation ProcedureA nn otation D ataB aseS eque n ce fro m 25W h ole gen om e sC o m p ared w ith C O GS elect co ntig fo rO R F pre dictionS eque n ce colle

11、ctio nS eque n ce F un ctio n In dexS eque n ce C lassficationP erl P ro cessin g M o d uleR esult F ile P arsin gC re at F u n ction C lassF o r A ceD B D ataB aseR ep ort F ileP rintin g P S I - B lastSequence Data Storage and DistributionDatabasePerl modulesAceclient/perlGifaceserverCacheDb.xxx.h

12、tmlDb.xxx.keysetDb.xxx.boxesDb.xxx.gif Cgi-binwebacedisplaygetimglabelHtml,gifJavaScriptJavaCookiesINTERNETData Analysis and Model DesignA. Complete Genome Analysis大規(guī)模全基因組shotgun測(cè)序Shotgun 反應(yīng): 109,970有效總讀長(zhǎng)：26,553,803bp基因組總長(zhǎng)：2,689,445bp冗余度：9.87PCR補(bǔ)洞PCR反應(yīng)：40,000PCR產(chǎn)物測(cè)序：4,100Circular representation of t

13、he genome of T. tengcongensis MB4Sequence Data DistributionVectorSequencing and Data Processing Procedure BasecallingRepeat MarkORF Prediction Gene AnnotationFinishingAssembleVector MarkPhredPhd2fastaCrossmatchPhrapConsedRepeatmaskerGlimmerBlastxBlastnClastaltRNAscanQualityControlQualCalQualDrawQual

14、StatCOGsSwiss-port,PIR, GDB,GenBankSequencingComplexity analysisICCCffNNfNlLlllLl220011()()The original presentation of coincident index is:in which N is the number of characters in the sequence,L is the number of different kinds of character, and is the frequency of the lth character appearing in t

15、his sequence.flThe coincident index introduced from cryptology is applied to DNA sequence analysis. In case of mono-letter,the expressing of coincident index IC of DNA sequence is:ICSTPILbbRIPjkjkkj()()1641j=0,1,.,L-1; k=A,C,G,T.R=4 25 (25 1)=2400 In case of biletter,the expression of IC is:ICSTPILb

16、bRIPjkjkkj()()1641j=0,1,.,L-1; k=AA,AC,AG,AT,CA,CC,.TG,TT R1 61 0 01 61 0 01 615 2 5()Auto-annotation ProcedureA nn otation D ataB aseS eque n ce fro m 25W h ole gen om e sC o m p ared w ith C O GS elect co ntig fo rO R F pre dictionS eque n ce colle ctio nS eque n ce F un ctio n In dexS eque n ce C lassfica