1、31 1 2001 M1 Ll MMATHEMATICS IN PRACTICE AND THEORYVol. 31 No. 1 Jan. 2001 sequences. Thesecond is the periodic property of the DN A sequences. The third is that amountof information of the sequences. By using this method, we classify the nature sequences andartifical sequences. At last, we analyze
2、the characteristic in this model and consider thegeneralization of this model.1DNA s 5 冯 涛, 康吉吉雯, 韩小军指导老师: 贺明峰(v v,v 116024)I: d9ZE4 |“+,TBP * ,M ATLABM E. %s 5, T.4 |+ H I n ,N$ * ZE,? * adL5、 1 a ? . b#,3e .K1: 4 B| * DNAs ZE.T n5 qd9ZE20X Y DNA +4 |,DNA +_ ,i|T“ BP * .T M ATLAB qNeural Network To
3、olbox( * Q)Q_.l(Backpropagation BP) E * .,T/ BP * ,|4 |DNA+_“T“sY .YV ,|20s “1821 “4 |+_ i s .TV :4s ZE ? qDNA s ,| * DNA s V.1 5 ( )DNA aA、 T、 C、 GB? p 7.X 1- 10 A ,11- 20 B .5, 1T :1)4 |A、 B +;2) 4 |A、 B +G ,20 #1821 sA、 B ( V ?i H A、 B +, ?BA、 B B ).5,5 1- 20G ,4 |A、 B d9+, *BP s MY.2 y G * M ?ZB
4、v? is) dL“d1, 1+:1) ? i/ i dLf ;2) 5 1 a1 ? ;3) - V“.4.4 #_X Y 1 20, |A -7 (1 7)B -7 (11 17)T “Strain, 8 10、 18 20T_“Stest 1: 25- 5- 1# 19- 5- 1 ,“98S10- 5. l S,_“s _, I s MY . 1、 2 S (- 0. 2 +0. 2W . E n MBP E,M MMZE,4 l . 1 7 S H, q |0. 9( 2 |1. 0),8“ T |0. 6( 2 |T0. 7) , T |10. “Eh , / +,N H,vh Z
5、#T,9vT, l S. r H, “ q wL m1(a)m2(a) ( ),N H_“_T m1(b)m2(b)( ):m1(a)m2(a), 1 303, 2 241,r 1 p, ( q y,r q .m1(b)m2(b), T 10% , * N H 1_“s 98. 3% , 294. 7% , q (100% , |_“F “,F“Strain+ test. N“. l , 1、2 V s MY .5 T#s5.1 21 40s M ATLAB q * Q( BP ) s .?C: 0. 9- 0. 9Ts S, BP qM, f , 1 . f Cy : 2 M 1 ,“ M
6、f /,y s;31F T 31 ,yN,+ 31 ? ?z8C M1.VQ 、_、s , ?C: 1 2 y, us,yN, 1. , 1s T,: A : 22,23, 25, 27, 29, 34, 35, 37, 39; B : 21,24,26,28,30, 31,32,33, 36,38,40.5.2 1821 s 21 40X s F “, 1 ,r10- 5.sY0, 0. 2, 0. 50, 0. 2, 0. 5Ts S,1821 s T: ( )“s SM,s q M.0Ts S V1821 s 7.291 !:1DNA s 5 6 #Z_:y+dL“d 4 ZVr ,7
7、O V “K,.ds MYZEAr, * V z ? , Q ,4 . ID:1 , y. * e., 1998.2 Funahashi K J. the Appronimate Realization of Continuous Mapping by Neural Networks. Neutral Networks, 1989,(2).3 Rumelhart D E, Moclell J L. Parallel Distributed Processing: Exploration in the Microstructure of cognition. M ITPress, London,
8、 1996.4 ; . * # . bv, 1999.5 . * .v v, 1995.6 L ?, .M ATLAB“ds !9 * .0 S/v, 1999.7 V,t?.5 p ? * ZE. , 1995.8 o . * / .S S/v, 1993.AModelforDNASequenceClusteringProblemFENG Tao, KANG Zhe-wen, HAN Xiao-jun(Dalian University of Technology, Dalian 116024)Abstract: This paper presents a method applying a
9、rtificial neural network ( NN ) to DN Aclustering problem. First we use theprobability statistics method to extract thecharacters fromthe 20 artificial DNA sequences whose categories are known. Thus we can get the charactervectors of the DN A sequences and input them as samples into BP neuron NN for
10、 learning. We30 L l M31 31 1 2001 M1 Ll MMATHEMATICS IN PRACTICE AND THEORYVol. 31 No. 1 Jan. 2001 employ the BP ( back propagation) algorithm to train NN by use of the Neural NetworkToolbox in M ATLAB software package. In this paper, two three-story NN are created to inputthe extracted DN A charact
11、er vectors as samples into them. After the training, characters areextracted from the 20 unclassified artificial sequence samples and 182 natural sequence samplesto form the character vectors as input of the two NN for clustering. The results shows: theclustering method presented in this paper can c
12、lassify the DNA sequences in quite high accuracyand precision. It is quite feasible to apply the artificial neural network to DNA sequenceclustering.DNA s 杨 健, 王 驰, 杨 勇指导老师: 王 鸣(v, 100871)I: |DNA aF AT“c”1oM,E1oMiss ZEs .V ,BZE V4 | z+,7 Os 9 %.B B7/,s51 %7L L=, b#, Y e .K1: s +,d9、K ZE,VX“ 4 ? zV +
13、1o31, N4 s S, ? z DNA s . n5,VX“ 5E C31,i9 S q#s . , “ Kl=E 3115)f ,i O Pr,4 V.K, 5)f s1oM, ,QsE“s ,iA ,V7 B1 Ys ZE.V_,NZE21 40|“ zs , 182DNA “ T,9 zrT.1 5 ( )2 L !(1)Ls “21 40; A 9 B “s1V5% .(2)L !keyword1ts 1“, Ots1( t,sln/).(3)L !VA、 B +31DNA ( s.3 sV DNA 4?C,31C q,7 Ot31AB CQ A , U t31TA, B Bs S
14、. A、 B X“Sd9s,s V ?V +31.y311 V ?B“, t311 ,K ?V +Bs31. t31TS A,B , A S ,K BDNA s .31 1 2001 M1 Ll MMATHEMATICS IN PRACTICE AND THEORYVol. 31 No. 1 Jan. 2001 DNA s 汤诗杰, 周 亮, 王晓玲指导老师: 孙广中(S S/v,g 230026)I: 4 DNA s , B,A、 G、 T、 C aC q; = Ba W,B B I n q ?c; ,TDNA jB , I n 9FB3 9 .5 lZ T V) , X, +).Ks ZE
15、, 8 P.K1: DNA s L=5,4 M . z8CDNA +, s ZE S, BKHq f /, ? sQ Z +.G 4 YS,BSs E 1 p. ZE ZE8“. ZEsY8C C q, C , c . ZE, #1 s T.K,s s,i wT ) .1 54( )2 5s B1 s 5, V ZL, ZE 5.X3 S1, S2, S3 S40, Si= x1x2x3 xni , xj a,t, c,g ;3 “A, B, A B= O,i1i10 H, Si A;11i20 H, Si B.C1 p I n21i40 H, Si“A#“B1“. ,51o 1VXsz 203 4 |s +. t+, V1 *tS s ./ |n5s S5A1) .3 s S# Nn5, 4 |+ / Hq:(1) |+A VSAFBF.9 , t+ Vz usXS Us 20 . 1 A B .(2) |+A B L=il.B % ?$ j.1 , T In L=il, V 7h3 s S:XB E gt 7 S,7XA 10 gt 7 S, .g 7 S .A BHq. TyN + 1,ie +| s s ,A . .