A computational frame and resource for understanding the lncRNA-environmental factor associations and prediction of environmental factors implicated in diseases

Abstract

The complex traits of an organism are associated with a complex interplay between genetic factors (GFs) and environmental factors (EFs). However, compared with protein-coding genes and microRNAs, there is a paucity of computational methods and bioinformatic resource platform for understanding the associations between lncRNA and EF. In this study, we developed a novel computational method to identify potential associations between lncRNA and EF, and released LncEnvironmentDB, a user-friendly web-based database aiming to provide a comprehensive resource platform for lncRNA and EF. Topological analysis of EF-related networks revealed the small world, scale-free and modularity structure. We also found that lncRNA and EF significantly enriched interacting miRNAs are functionally more related by analyzing their related diseases, implying that the predicted lncRNA signature of EF can reflect the functional characteristics to some degree. Finally, we developed a random walk with a restart-based computational model (RWREFD) to predict potential disease-related EFs by integrating lncRNA-EF associations and EF–disease associations. The performance of RWREFD was evaluated by experimentally verified EF–disease associations based on leave-one-out cross-validation and achieved an AUC value of 0.71, which is higher than randomization test, indicating that the RWREFD method has a reliable and high accuracy of prediction. To the best of our knowledge, LncEnvironmentDB is the first attempt to predict and house the experimental and predicted associations between lncRNA and EF. LncEnvironmentDB is freely available on the web at http://bioinfo.hrbmu.edu.cn/lncefdb/.