Association Analysis Between Long Non-coding RNAs And Diseases

Bioinformatics is a newly arisen science which is comprised of biology, computer science and applied mathematics. People choose an appropriate computing model to process data based on biological experimental acquired. The results of data retrieval and analysis reveals the potential biological significance of these data. Research on biological data analysis can advance the medical science, and provides reference data for future clinical trials, which has a very important theoretical value and practical significance.Only 1% to 2% of the human genome is in charge of protein-coding genes, the rest of the non-coding region earlier is considered to be useless. Increasing evidence from public studies has demonstrated their critical roles in various biological processes through a variety of mechanisms such as Genomic imprinting,Cell differentiation variation, immune response, the process of tumorigenesis and the rest of the physical.According to the existing human lncRNA-disease association data. Although several computational methods have been developed to predict disease-related lncRNA, this still remains a considerable challenging task. As expected, functional analysis showed that lnc RNA sharing significantly enriched interacting mi RNAs tend to be involved in similar diseases and have more functionally related flanking gene sets.For the prediction of lnc RNA-diseases association,this paper analyzes the inferring novel lnc RNA-disease associations based on random walk on lncRNA founctional similarity network, and theK-Nearest Neighbor method. we propose a new algorithm to infer the correlation of lncRNA-diseases.The main research can be described as follows:1. A lncRNA-Disease prediction method is proposed based on heterogeneous networks, which contains lnc RNA similarity network, disease similarity network and lncRNA-disease assisoations.Moreover these similarity matrix are improved. Then a random walk algorithm is adopted to predict lnc RNA-disease association. Leave-one-out cross-validation experiment shows that AUC equals to0.9544,which is better than strategy based on lncRNA functional similarity network.2. A Lnc RNA-Disease associations prediction based on double K-Nearest Neighbour is presented which adopted an extended K-nearest neighbor algorithm. The improved method can avoid the redundary problem existing in original K-Nearest Neighbour. The leave-one-out cross-validation experiment shows,AUC equals to 0.8685, which is better than the traditional K-Nearest Neighbour.3. A lnc RNA-disease online inquiry webset based on B / S architecture is designed and implemented.To support the development of network medicine for lncRNA-disease association discovery., we build an online website based on B / S architecture, using.NET + SQL data model. The key modules contains information searching, data download and data updating. Software tests shows that the system has good stability.