Small non-coding RNAs (e.g. miRNAs) and long non-coding RNAs (e.g. lincRNAs and circRNAs) are emerging as key regulators of various cellular processes. However, only a very small fraction of these enigmatic RNAs have been well functionally characterized. In this study, we describe deepBase v2.0 (http://biocenter.sysu.edu.cn/deepBase/), an updated platform, to decode evolution, expression patterns and functions of diverse ncRNAs across 19 species. deepBase v2.0 has been updated to provide the most comprehensive collection of ncRNA-derived small RNAs generated from 588 sRNA-Seq datasets. Moreover, we developed a pipeline named lncSeeker to identify 176 680 high-confidence lncRNAs from 14 species. Temporal and spatial expression patterns of various ncRNAs were profiled. We identified approximately 24 280 primate-specific, 5193 rodent-specific lncRNAs, and 55 highly conserved lncRNA orthologs between human and zebrafish. We annotated 14 867 human circRNAs, 1260 of which are orthologous to mouse circRNAs. By combining expression profiles and functional genomic annotations, we developed lncFunction web-server to predict the function of lncRNAs based on protein-lncRNA co-expression networks. This study is expected to provide considerable resources to facilitate future experimental studies and to uncover ncRNA functions.

TRNA-derived small RNA fragments (tRFs) are one class of small non-coding RNAs derived from transfer RNAs (tRNAs). tRFs play important roles in cellular processes and are involved in multiple cancers. High-throughput small RNA (sRNA) sequencing experiments can detect all the cellular expressed sRNAs, including tRFs. However, distinguishing genuine tRFs from RNA fragments generated by random degradation remains a major challenge. In this study, we developed an integrated web-based computing system, tRF2Cancer, to accurately identify tRFs from sRNA deep-sequencing data and evaluate their expression in multiple cancers. The binomial test was introduced to evaluate whether reads from a small RNA-seq data set represent tRFs or degraded fragments. A classification method was then used to annotate the types of tRFs based on their sites of origin in pre-tRNA or mature tRNA. We applied the pipeline to analyze 10 991 data sets from 32 types of cancers and identified thousands of expressed tRFs. A tool called ‘tRFinCancer’ was developed to facilitate the users to inspect the expression of tRFs across different types of cancers. Another tool called ‘tRFBrowser’ shows both the sites of origin and the distribution of chemical modification sites in tRFs on their source tRNA. The tRF2Cancer web server is available at http://rna.sysu.edu.cn/tRFfinder/.

Non-coding RNAs (ncRNAs) are emerging as key regulators of various biological processes. Although thousands of ncRNAs have been discovered, the transcriptional mechanisms and networks of the majority of ncRNAs have not been fully investigated. In this study, we updated ChIPBase to version 3.0 (https://rnasysu.com/chipbase3/) to provide the most comprehensive transcriptional regulation atlas of ncRNAs and protein-coding genes (PCGs). ChIPBase has identified ∼151 187 000 regulatory relationships between ∼171 600 genes and ∼3000 regulators by analyzing ∼55 000 ChIP-seq datasets, which represent a 30-fold expansion. Moreover, we de novo identified ∼29 000 motif matrices of transcription factors. In addition, we constructed a novel ‘Enhancer’ module to predict ∼1 837 200 regulation regions functioning as poised, active or super enhancers under ∼1300 conditions. Importantly, we constructed exhaustive coexpression maps between regulators and their target genes by integrating expression profiles of ∼65 000 normal and ∼15 000 tumor samples. We built a ‘Disease’ module to obtain an atlas of the disease-associated variations in the regulation regions of genes. We also constructed an ‘EpiInter’ module to explore potential interactions between epitranscriptome and epigenome. Finally, we designed ‘Network’ module to provide extensive and gene-centred regulatory networks. ChIPBase will serve as a useful resource to facilitate integrative explorations and expand our understanding of transcriptional regulation.

Relapse is a major challenge in the successful treatment of childhood acute lymphoblastic leukemia (ALL). Despite intensive research efforts, the mechanisms of ALL relapse are still not fully understood. An understanding of the molecular mechanisms underlying treatment outcome, therapy response and the biology of relapse is required. In this study, we carried out a genome-wide microRNA (miRNA) microarray analysis to determine the miRNA expression profiles and relapse-associated miRNA patterns in a panel of matched diagnosis-relapse or diagnosis-complete remission (CR) childhood ALL samples. A set of miRNAs differentially expressed either in relapsed patients or at diagnosis compared with CR was further validated by quantitative real-time polymerase chain reaction in an independent sample set. Analysis of the predicted functions of target genes based on gene ontology 'biological process' categories revealed that the abnormally expressed miRNAs are associated with oncogenesis, classical multidrug resistance pathways and leukemic stem cell self-renewal and differentiation pathways. Several targets of the miRNAs associated with ALL relapse were experimentally validated, including FOXO3, BMI1 and E2F1. We further investigated the association of these dysregulated miRNAs with clinical outcome and confirmed significant associations for miR-708, miR-223 and miR-27a with individual relapse-free survival. Notably, miR-708 was also found to be associated with the in vivo glucocorticoid therapy response and with disease risk stratification. These miRNAs and their targets might be used to optimize anti-leukemic therapy, and serve as novel targets for development of new countermeasures of leukemia. This fundamental study may also contribute to establish the mechanisms of relapse in other cancers.

Although over 170 chemical modifications have been identified, their prevalence, mechanism and function remain largely unknown. To enable integrated analysis of diverse RNA modification profiles, we have developed RMBase v3.0 (http://bioinformaticsscience.cn/rmbase/), a comprehensive platform consisting of eight modules. These modules facilitate the exploration of transcriptome-wide landscape, biogenesis, interactome and functions of RNA modifications. By mining thousands of epitranscriptome datasets with novel pipelines, the ‘RNA Modifications’ module reveals the map of 73 RNA modifications of 62 species. the ‘Genes’ module allows to retrieve RNA modification profiles and clusters by gene and transcript. The ‘Mechanisms’ module explores 23 382 enzyme-catalyzed or snoRNA-guided modified sites to elucidate their biogenesis mechanisms. The ‘Co-localization’ module systematically formulates potential correlations between 14 histone modifications and 6 RNA modifications in various cell-lines. The ‘RMP’ module investigates the differential expression profiles of 146 RNA-modifying proteins (RMPs) in 18 types of cancers. The ‘Interactome’ integrates the interactional relationships between 73 RNA modifications with RBP binding events, miRNA targets and SNPs. The ‘Motif’ illuminates the enriched motifs for 11 types of RNA modifications identified from epitranscriptome datasets. The ‘Tools’ introduces a novel web-based ‘modGeneTool’ for annotating modifications. Overall, RMBase v3.0 provides various resources and tools for studying RNA modifications.

Long non-coding RNAs (lncRNAs) are emerging as important regulatory molecules in developmental, physiological and pathological processes. However, the precise mechanism and functions of most of lncRNAs remain largely unknown. Recent advances in high-throughput sequencing of immunoprecipitated RNAs after cross-linking (CLIP-Seq) provide powerful ways to identify biologically relevant protein–lncRNA interactions. In this study, by analyzing millions of RNA-binding protein (RBP) binding sites from 117 CLIP-Seq datasets generated by 50 independent studies, we identified 22,735 RBP-lncRNA regulatory relationships. We found that one single lncRNA will generally be bound and regulated by one or multiple RBPs, the combination of which may coordinately regulate gene expression. We also revealed the expression correlation of these interaction networks by mining expression profiles of over 6000 normal and tumor samples from 14 cancer types. Our combined analysis of CLIP-Seq data and genome-wide association studies (GWAS) data discovered hundreds of disease-related SNPs resided in the RBP binding sites of lncRNAs. Finally, we developed interactive web implementations to provide visualization, analysis and downloading of the aforementioned large-scale datasets. Our study represented an important step in identification and analysis of RBP-lncRNA interactions and showed that these interactions may play crucial roles in cancer and genetic diseases.