Font Size: a A A

The Design And Implementation Of The Functional Analysis System Of Isoforms

Posted on:2017-10-11Degree:MasterType:Thesis
Country:ChinaCandidate:X A WuFull Text:PDF
GTID:2348330566456742Subject:Software engineering
Abstract/Summary:PDF Full Text Request
The human gene expression,mostly produce a variety of different functions of isoforms by alternative splicing.For diversity of the genetic function,selective transcription process is a very important mechanism.In recent years,more and more biological experiment,especially the biological gene protein tests are involving the selective expression of gene.there are more and more experiments that study biological information began from the isoforms level,analyze the information contained in biological gene.however,Different from genes,the current understanding of these numerous isoforms is less,lack a isoforms function data integration library like gene function database.Along with the increase of the transcriptome sequencing data,there are some methods such as a method based on the structure of 3 d conformation,and the expectation maximization algorithm.But these methods are not mature enough or stay on the algorithm research stage,and don`t have the actual experiment or mature system.Under such background,we use the machine learning methods,integrate the sequencing data.this paper will study to establish a analysis system of the isoforms function.This article divide the function analysis system of isoforms into data acquisition and preconditioning system,the building of prediction data,application of prediction methods,prediction results show part etc.the main data source of System is NCBI RNA-seq database and the Gene Ontology database,using the sequence alignment tool and express value estimate tool to estimate express quantity for expression spectrum.We use the mapping relationship between GO and Gene and the mapping relationship between genes and isoforms for functional annotation of the training set,the all data will be function prediction data sources.the multi-instance label propagation algorithm oriented isoforms function prediction method and the framework based on multiple instance multiple label isoforms function prediction method is applied to the analysis of the isoforms function.At first,this paper describes the research background and significance,isoforms and summarizing the previous research and the system knowledge,the overall architecture of the system,and then we introduce the contents of each part based on system structure,and show the research results.finally we show the vision of thefuture of our work.
Keywords/Search Tags:Biological information, isoforms, functional annotation, machine learning
PDF Full Text Request
Related items