Dynamic Programming Sequence alignment：动态规划序列比对ppt课件.ppt

《Dynamic Programming Sequence alignment：动态规划序列比对ppt课件.ppt》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、DNASequenceComparison:FirstSuccessStoryFindingsequencesimilaritieswithgenesofknownfunctionisacommonapproachtoinferanewlysequencedgene’sfunctionIn1984RussellDoolittleandcolleaguesfoundsimilaritiesbetweencancer-causinggeneandnormalgrowthfactor(PDGF)geneAnormalgrowthgeneswitchedonatthewrongtimeca

2、usescancer!CysticFibrosisCysticfibrosis(CF)isachronicandfrequentlyfatalgeneticdiseaseofthebody'smucusglands.CFprimarilyaffectstherespiratorysystemsinchildren.SearchfortheCFgenewasnarrowedto~1Mbp,andtheregionwassequenced.Scannedadatabaseformatchestoknowngenes.Asegmentinthisregionmatchedthegenef

3、orsomeATPbindingprotein(s).Theseproteinsarepartoftheiontransportchannel,andCFinvolvessweatsecretionswithabnormalsodiumcontent!RoleforBioinformaticsGenesimilaritiesbetweentwogeneswithknownandunknownfunctionalertbiologiststosomepossibilitiesComputingasimilarityscorebetweentwogenestellshowlikelyi

4、tisthattheyhavesimilarfunctionsDynamicprogrammingisatechniqueforrevealingsimilaritiesbetweengenesMotivatingDynamicProgrammingDynamicprogrammingexample:ManhattanTouristProblemImagineseekingapath(fromsourcetosink)totravel(onlyeastwardandsouthward)withthemostnumberofattractions(*)intheManhattang

5、ridSink***********Source*Imagineseekingapath(fromsourcetosink)totravel(onlyeastwardandsouthward)withthemostnumberofattractions(*)intheManhattangridSink***********Source*Dynamicprogrammingexample:ManhattanTouristProblemManhattanTouristProblem:FormulationGoal:Findthelongestpathinaweightedgrid.I

6、nput:AweightedgridGwithtwodistinctvertices,onelabeled“source”andtheotherlabeled“sink”Output:AlongestpathinGfrom“source”to“sink”MTP:AnExample32407333013244564655822501230123jcoordinateicoordinate13sourcesink432401024331122241995152302034MTP:GreedyAlgorithmIsNotOptimal1252152340005303501035512pr

7、omisingstart,butleadstobadchoices!sourcesink1822MTP:SimpleRecursiveProgramMT(n,m)ifn=0orm=0returnMT(n,m)xMT(n-1,m)+lengthoftheedgefrom(n-1,m)to(n,m)yMT(n,m-1)+lengthoftheedgefrom(n,m-1)to(n,m)returnmax{x,y}What’swrongwiththisapproach?