基于蛋白質(zhì)序列信息預測二硫鍵的新方法研究.pdf

1、聲明本學位論文是我在導師的指導下取得的研究成果，盡我所知，在本學位論文中，除了加以標注和致謝的部分外，不包含其他人已經(jīng)發(fā)表或公布過的研究成果，也不包含我為獲得任何教育機構(gòu)的學位或?qū)W歷而使用過的材料。與我一同工作的同事對本學位論文做出的貢獻均己在論文中作了明確的說明。研究生簽名：塞避kf7年3月／婦學位論文使用授權(quán)聲明南京理工大學有權(quán)保存本學位論文的電子和紙質(zhì)文檔，可以借閱或上網(wǎng)公布本學位論文的部分或全部內(nèi)容，可以向有關(guān)部門或機構(gòu)送交并授

2、權(quán)其保存、借閱或上網(wǎng)公布本學位論文的部分或全部內(nèi)容。對于保密論文，按保密的有關(guān)規(guī)定和程序處理。研究生簽名：乙)7年3月弓日Abastract碩士學位論文AbstractProteinisthematerialbasisofalllifeactivitiesanditplaysanimportantroleinlifephenomenonandphysiologicalfunctionProteomicsisanimportantbran

3、chbelongingtobioinformaticsByresearchingofthisbranch，humansCanunderstandthenatureoflifeprocessbetter，andCangainthetheoreticalfoundationofthepreventionandtreatmentofdiseases，andasaresult，OurlivingqualityCanalsobeimprovedI

4、nrecentyears，governmentsandresearchersfromallovertheworldattachesgreatimportancetothissubjectWiththecomingofpostgenomics，ProteomicshasbecomethemainresearchdirectionofbioinformaticsProteinconsistsofhugeofAminoacidresidues

5、viacomplexchemicalreactions，containingcomplexstructuresandfunctions，whichfunctionsarerealizedbythousandsofcombinations誦thinteractionsamong20differentaminoacidsDisulfidebondplaysanimportantroleinmaintainingtheconstructsan

6、dfunctionsofprotein，whosebondingstateandconnectivitypattemdeserveextensiveresearchInrecentyears，withtherapiddevelopmentofproteinsequencingtechnology，thenumberofknownproteinsequencesaswellasthedisulfidebondsisgrowingexplo

7、sivelyUsingtraditionalmethodsbasedonBiologytestingdisulfidebondnotonlycostsmuch，butalsospendsmuchtimeSo，predictingdisulfidebondaccuratelyandfastbycomingupintelligentalgorithmsisthepriorityInrecentyears，differentkindsofal

8、gorithmsbasedonMachineLearninghavebeenusedondisulfidebondpredictionCysteinepairsinaproteinsequencehavetwostates(theycanformadisulfidebondornot)，whichbelongstoatypicalbinaryclassificationproblemTheformerresearcherexperime

9、ntallyaddressedtheproblemofdisulfidebondpredictionthroughfeatureextractionandfeatureselection，andthengainfinalresultsthroughclassifierWhenresearchingtheconnectivitypatterns，thispaperdividedthefeaturesintoaminoacidresidue

10、sfeatures，cysteinefeaturesandglobalfeaturestofindthemosteffectivefeaturegroupIntheprocessoffeatureselection，firstly，calculatingeachfeature’Sscore(includingvariancescore，LaplacianscoreandFisherscore)tochooserepresentative

11、features，thencarryingoutcorrelationfeatureselectionFinally，predictingconnectivitypatternsthroughdifferentclassifiersComparingwithformerresearch，thispapercardedoutcorrelationfeatureselectionThealgorithmandbenchmarkdataset