Research On Differential Analysis Algorithm And Software Platform Development

With the completion of the mapping of genetic maps,the field of bioinformatics has begun to study the mechanisms of cell functions and complex diseases at the molecular level.In recent years,the rapid development of high-throughput sequenc-ing technology has produced maritime gene sequencing data,which makes it possible to study the mechanism of diseases.In addition,in the field of machine learning,researchers have proposed various differential analysis models and algorithms,which provide powerful theoretical tools and technical supports for identifying genes as-sociated with genes from genomics data.Based on these studies,the following two researches are carried out:(1)A differential network construction algorithm based on interactive linear regression is proposed.(2)An R software package for differ-entially expressed gene identifying for single-cell sequencing data is developed.The main contributions are as follows:1.In order to integrate genomic data from different platforms and determine the key genes leading to gene network perturbation,this paper proposes a differ-ential network construction model which is based on interactive linear regression model.The model uses group sparse regularization to explore the commonality between different data platforms and to mine important gene which may drive the disturbance of gene network.We also propose an alternating direction multiplier algorithm to solve the model.Simulation experiments have shown the effectiveness of the proposed model.This model is also applied to identify the gene network rewiring between the pro-neuronal and mesenchymal subtypes of breast cancer.Ex-periment results indicate that the hub genes in the identified differential networks have important biological functions.2.This paper also develops an application software platform to integrate mul-tiple differential expression analysis algorithms designed for single-cell sequencing data.Based on the R language,the platform integrates the current popular differ-ential expression analysis algorithms.The main contributions include the following aspects:(1)A variety of algorithms are integrated with a unified input and output format,which facilitates the users to perform differential expression analysis and compare the performance of different algorithms.(2)Related downstream analysis functions(e.g.,Gene Ontology enrichment analysis and KEGG pathway enrichment analysis)are also implemented.(3)An R package and Shiny Application have been developed to provide a convenient GUI interface.