ImmunemiR

MicroRNAs are a class of small non-coding regulatory RNA molecules that modulate the expression of several genes at post-transcriptional level and play a vital role in disease pathogenesis. Recent research shows that a range of miRNAs are involved in the regulation of immunity and its deregulation results in immune mediated diseases such as cancer, inflammation and autoimmune diseases. Computational discovery of these immune miRNAs using a set of specific features is highly desirable. In the current investigation, we present a SVM based classification system which uses a set of novel network based topological and motif features in addition to the baseline sequential and structural features to predict immune specific miRNAs from other non-immune miRNAs. The classifier was trained and tested on a balanced set of equal number of positive and negative examples to show the discriminative power of our network features. Experimental results show that our approach achieves an accuracy of 90.2% and outperforms the classification accuracy of 63.2% reported using the traditional miRNA sequential and structural features. The proposed classifier was further validated with two immune disease sub-class datasets related to multiple sclerosis microarray data and psoriasis RNA-seq data with higher accuracy. These results indicate that our classifier which uses network and motif features along with sequential and structural features will lead to significant improvement in classifying immune miRNAs and hence can be applied to identify other specific classes of miRNAs as an extensible miRNA classification system.[1]

MicroRNAs are the key regulators of gene expression and their abnormal expression in the immune system may be associated with several human diseases such as inflammation, cancer and autoimmune diseases. Elucidation of microRNA (miRNA) disease association through the interactome will deepen the understanding of its disease mechanisms. In this present study, miRNAs specific to immune related diseases were retrieved from curated databases and literature based on MeSH classification of immune system diseases. In total 245 immune miRNAs associated with 92 OMIM disease categories were identified and they are prioritized to specific immune diseases using random walk ranking algorithm. These data were compiled as ImmunemiR, a database of prioritized immune miRNA disease associations. This database provides both text based annotation information and network visualization of its interactome network. To our knowledge, ImmunemiR is the first available database to provide a comprehensive repository of human immune disease associated miRNAs with network visualization options of its target genes, protein-protein interactions (PPI).[2]

References