Big data and deep learning for RNA biology.

The exponential growth of big data in RNA biology (RB) has led to the development of deep learning (DL) models that have driven crucial discoveries. As constantly evidenced by DL studies in other fields, the successful implementation of DL in RB depends heavily on the effective utilization of large-scale datasets from public databases. In achieving this goal, data encoding methods, learning algorithms, and techniques that align well with biological domain knowledge have played pivotal roles. In this review, we provide guiding principles for applying these DL concepts to various problems in RB by demonstrating successful examples and associated methodologies. We also discuss the remaining challenges in developing DL models for RB and suggest strategies to overcome these challenges. Overall, this review aims to illuminate the compelling potential of DL for RB and ways to apply this powerful technology to investigate the intriguing biology of RNA more effectively.

An interaction between eIF4A3 and eIF3g drives the internal initiation of translation.

An RNA structure or modified RNA sequences can provide a platform for ribosome loading and internal translation initiation. The functional significance of internal translation has recently been highlighted by the discovery that a subset of circular RNAs (circRNAs) is internally translated. However, the molecular mechanisms underlying the internal initiation of translation in circRNAs remain unclear. Here, we identify eIF3g (a subunit of eIF3 complex) as a binding partner of eIF4A3, a core component of the exon-junction complex (EJC) that is deposited onto spliced mRNAs and plays multiple roles in the regulation of gene expression. The direct interaction between eIF4A3-eIF3g serves as a molecular linker between the eIF4A3 and eIF3 complex, thereby facilitating internal ribosomal entry. Protein synthesis from in vitro-synthesized circRNA demonstrates eIF4A3-driven internal translation, which relies on the eIF4A3-eIF3g interaction. Furthermore, our transcriptome-wide analysis shows that efficient polysomal association of endogenous circRNAs requires eIF4A3. Notably, a subset of endogenous circRNAs can express a full-length intact protein, such as ß-catenin, in an eIF4A3-dependent manner. Collectively, our results expand the understanding of the protein-coding potential of the human transcriptome, including circRNAs.

AIVariant: a deep learning-based somatic variant detector for highly contaminated tumor samples.

The detection of somatic DNA variants in tumor samples with low tumor purity or sequencing depth remains a daunting challenge despite numerous attempts to address this problem. In this study, we constructed a substantially extended set of actual positive variants originating from a wide range of tumor purities and sequencing depths, as well as actual negative variants derived from sequencer-specific sequencing errors. A deep learning model named AIVariant, trained on this extended dataset, outperforms previously reported methods when tested under various tumor purities and sequencing depths, especially low tumor purity and sequencing depth.

Determinants of Functional MicroRNA Targeting.

MicroRNAs (miRNAs) play cardinal roles in regulating biological pathways and processes, resulting in significant physiological effects. To understand the complex regulatory network of miRNAs, previous studies have utilized massivescale datasets of miRNA targeting and attempted to computationally predict the functional targets of miRNAs. Many miRNA target prediction tools have been developed and are widely used by scientists from various fields of biology and medicine. Most of these tools consider seed pairing between miRNAs and their mRNA targets and additionally consider other determinants to improve prediction accuracy. However, these tools exhibit limited prediction accuracy and high false positive rates. The utilization of additional determinants, such as RNA modifications and RNA-binding protein binding sites, may further improve miRNA target prediction. In this review, we discuss the determinants of functional miRNA targeting that are currently used in miRNA target prediction and the potentially predictive but unappreciated determinants that may improve prediction accuracy.

Ultrafast prediction of somatic structural variations by filtering out reads matched to pan-genome k-mer sets.

Variant callers typically produce massive numbers of false positives for structural variations, such as cancer-relevant copy-number alterations and fusion genes resulting from genome rearrangements. Here we describe an ultrafast and accurate detector of somatic structural variations that reduces read-mapping costs by filtering out reads matched to pan-genome k-mer sets. The detector, which we named ETCHING (for efficient detection of chromosomal rearrangements and fusion genes), reduces the number of false positives by leveraging machine-learning classifiers trained with six breakend-related features (clipped-read count, split-reads count, supporting paired-end read count, average mapping quality, depth difference and total length of clipped bases). When benchmarked against six callers on reference cell-free DNA, validated biomarkers of structural variants, matched tumour and normal whole genomes, and tumour-only targeted sequencing datasets, ETCHING was 11-fold faster than the second-fastest structural-variant caller at comparable performance and memory use. The speed and accuracy of ETCHING may aid large-scale genome projects and facilitate practical implementations in precision medicine.

A database of 5305 healthy Korean individuals reveals genetic and clinical implications for an East Asian population.

Despite substantial advances in disease genetics, studies to date have largely focused on individuals of European descent. This limits further discoveries of novel functional genetic variants in other ethnic groups. To alleviate the paucity of East Asian population genome resources, we established the Korean Variant Archive 2 (KOVA 2), which is composed of 1896 whole-genome sequences and 3409 whole-exome sequences from healthy individuals of Korean ethnicity. This is the largest genome database from the ethnic Korean population to date, surpassing the 1909 Korean individuals deposited in gnomAD. The variants in KOVA 2 displayed all the known genetic features of those from previous genome databases, and we compiled data from Korean-specific runs of homozygosity, positively selected intervals, and structural variants. In doing so, we found loci, such as the loci of ADH1A/1B and UHRF1BP1, that are strongly selected in the Korean population relative to other East Asian populations. Our analysis of allele ages revealed a correlation between variant functionality and evolutionary age. The data can be browsed and downloaded from a public website ( https://www.kobic.re.kr/kova/ ). We anticipate that KOVA 2 will serve as a valuable resource for genetic studies involving East Asian populations.

The regulatory impact of RNA-binding proteins on microRNA targeting.

Argonaute is the primary mediator of metazoan miRNA targeting (MT). Among the currently identified >1,500 human RNA-binding proteins (RBPs), there are only a handful of RBPs known to enhance MT and several others reported to suppress MT, leaving the global impact of RBPs on MT elusive. In this study, we have systematically analyzed transcriptome-wide binding sites for 150 human RBPs and evaluated the quantitative effect of individual RBPs on MT efficacy. In contrast to previous studies, we show that most RBPs significantly affect MT and that all of those MT-regulating RBPs function as MT enhancers rather than suppressors, by making the local secondary structure of the target site accessible to Argonaute. Our findings illuminate the unappreciated regulatory impact of human RBPs on MT, and as these RBPs may play key roles in the gene regulatory network governed by metazoan miRNAs, MT should be understood in the context of co-regulating RBPs.

A high-resolution temporal atlas of the SARS-CoV-2 translatome and transcriptome.

COVID-19 is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which infected >200 million people resulting in >4 million deaths. However, temporal landscape of the SARS-CoV-2 translatome and its impact on the human genome remain unexplored. Here, we report a high-resolution atlas of the translatome and transcriptome of SARS-CoV-2 for various time points after infecting human cells. Intriguingly, substantial amount of SARS-CoV-2 translation initiates at a novel translation initiation site (TIS) located in the leader sequence, termed TIS-L. Since TIS-L is included in all the genomic and subgenomic RNAs, the SARS-CoV-2 translatome may be regulated by a sophisticated interplay between TIS-L and downstream TISs. TIS-L functions as a strong translation enhancer for ORF S, and as translation suppressors for most of the other ORFs. Our global temporal atlas provides compelling insight into unique regulation of the SARS-CoV-2 translatome and helps comprehensively evaluate its impact on the human genome.

Aicardi-Goutières syndrome-associated gene SAMHD1 preserves genome integrity by preventing R-loop formation at transcription-replication conflict regions.

The comorbid association of autoimmune diseases with cancers has been a major obstacle to successful anti-cancer treatment. Cancer survival rate decreases significantly in patients with preexisting autoimmunity. However, to date, the molecular and cellular profiles of such comorbidities are poorly understood. We used Aicardi-Goutières syndrome (AGS) as a model autoimmune disease and explored the underlying mechanisms of genome instability in AGS-associated-gene-deficient patient cells. We found that R-loops are highly enriched at transcription-replication conflict regions of the genome in fibroblast of patients bearing SAMHD1 mutation, which is the AGS-associated-gene mutation most frequently reported with tumor and malignancies. In SAMHD1-depleted cells, R-loops accumulated with the concomitant activation of DNA damage responses. Removal of R-loops in SAMHD1 deficiency reduced cellular responses to genome instability. Furthermore, downregulation of SAMHD1 expression is associated with various types of cancer and poor survival rate. Our findings suggest that SAMHD1 functions as a tumor suppressor by resolving R-loops, and thus, SAMHD1 and R-loop may be novel diagnostic markers and targets for patient stratification in anti-cancer therapy.

Impact Ionization Induced by Accelerated Photoelectrons for Wide-Range and Highly Sensitive Detection of Volatile Organic Compounds at Room Temperature.

Ionization-based volatile organic compound (VOC) sensors that use photons or electrons operating at room temperature have attracted considerable attention as a promising alternative to conventional metal oxide-based sensors that require high temperature for sensing function. However, the photoionization sensors cannot ionize many gas species for their limited photon energy, and field emission-based ionization sensors that rely on the breakdown voltage of specific gas species in a pure state may not tell different concentration. This work demonstrates the detection of VOCs using impact ionization induced by accelerated photoelectrons. Although the photoelectrons emitted by relatively low photon energy typically have insufficient kinetic energy to cause impact ionization, in this approach, they are accelerated between microgap electrodes to enhance their kinetic energy such that the impact ionization of VOCs can be achieved. The demonstrated gas sensor sensitively detects toluene concentration in a wide range from 1000 ppm to 100 ppb with fast response and recovery time at room temperature. Additionally, diverse VOC species including benzene, p-xylene, and even acetylene with high ionization energy can be detected. The proposed method could be a viable solution for VOC sensors with low cost, scalable producibility, and high performance.

Identification of tumor suppressor miRNAs by integrative miRNA and mRNA sequencing of matched tumor-normal samples in lung adenocarcinoma.

The roles of miRNAs in lung cancer have not yet been explored systematically at the genome scale despite their important regulatory functions. Here, we report an integrative analysis of miRNA and mRNA sequencing data for matched tumor-normal samples from 109 Korean female patients with non-small-cell lung adenocarcinoma (LUAD). We produced miRNA sequencing (miRNA-Seq) and RNA-Seq data for 48 patients and RNA-Seq data for 61 additional patients. Subsequent differential expression analysis with stringent criteria yielded 44 miRNAs and 2322 genes. Integrative gene set analysis of the differentially expressed miRNAs and genes using miRNA-target information revealed several regulatory processes related to the cell cycle that were targeted by tumor suppressor miRNAs (TSmiR). We performed colony formation assays in A549 and NCI-H460 cell lines to test the tumor-suppressive activity of downregulated miRNAs in cancer and identified 7 novel TSmiRs (miR-144-5p, miR-218-1-3p, miR-223-3p, miR-27a-5p, miR-30a-3p, miR-30c-2-3p, miR-338-5p). Two miRNAs, miR-30a-3p and miR-30c-2-3p, showed differential survival characteristics in the Tumor Cancer Genome Atlas (TCGA) LUAD patient cohort indicating their prognostic value. Finally, we identified a network cluster of miRNAs and target genes that could be responsible for cell cycle regulation. Our study not only provides a dataset of miRNA as well as mRNA sequencing from the matched tumor-normal samples, but also reports several novel TSmiRs that could potentially be developed into prognostic biomarkers or therapeutic RNA drugs.

Interpretation of EBV infection in pan-cancer genome considering viral life cycle: LiEB (Life cycle of Epstein-Barr virus).

We report a novel transcriptomic analysis workflow called LiEB (Life cycle of Epstein-Barr virus) to characterize distributions of oncogenic virus, Epstein-Barr virus (EBV) infection in human tumors. We analyzed 851 The Cancer Genome Atlas whole-transcriptome sequencing (WTS) data to investigate EBV infection by life cycle information using three-step LiEB workflow: 1) characterize virus infection generally; 2) align transcriptome sequences against a hybrid human-EBV genome, and 3) quantify EBV gene expression. Our results agreed with EBV infection status of public cell line data. Analysis in stomach adenocarcinoma identified EBV-positive cases involving PIK3CA mutations and/or CDKN2A silencing with biologically more determination, compared to previous reports. In this study, we found that a small number of colorectal adenocarcinoma cases involved with EBV lytic gene expression. Expression of EBV lytic genes was also observed in 3% of external colon cancer cohort upon WTS analysis. Gene set enrichment analysis showed elevated expression of genes related to E2F targeting and interferon-gamma responses in EBV-associated tumors. Finally, we suggest that interpretation of EBV life cycle is essential when analyzing its infection in tumors, and LiEB provides high capability of detecting EBV-positive tumors. Observation of EBV lytic gene expression in a subset of colon cancers warrants further research.

Comparison of INDEL Calling Tools with Simulation Data and Real Short-Read Data.

Insertions and deletions (INDELs) comprise a significant proportion of human genetic variation, and recent papers have revealed that many human diseases may be attributable to INDELs. With the development of next-generation sequencing (NGS) technology, many statistical/computational tools have been developed for calling INDELs. However, there are differences among those tools, and comparisons among them have been limited. In order to better understand these inter-tool differences, five popular and publicly available INDEL calling tools-GATK HaplotypeCaller, Platypus, VarScan2, Scalpel, and GotCloud-were evaluated using simulation data, 1000 Genomes Project data, and family-based sequencing data. The accuracy of INDEL calling by each tool was mainly evaluated by concordance rates. Family-based sequencing data, which consisted of 49 individuals from eight Korean families, were used to calculate Mendelian error rates. Our comparison results show that GATK HaplotypeCaller usually performs the best and that joint calling with Platypus can lead to additional improvements in accuracy. The result of this study provides important information regarding future directions for the variant detection and the algorithms development.

Rules for functional microRNA targeting.

MicroRNAs (miRNAs) are ∼22nt-long single-stranded RNA molecules that form a RNA-induced silencing complex with Argonaute (AGO) protein to post-transcriptionally downregulate their target messenger RNAs (mRNAs). To understand the regulatory mechanisms of miRNA, discovering the underlying functional rules for how miRNAs recognize and repress their target mRNAs is of utmost importance. To determine functional miRNA targeting rules, previous studies extensively utilized various methods including high-throughput biochemical assays and bioinformatics analyses. However, targeting rules reported in one study often fail to be reproduced in other studies and therefore the general rules for functional miRNA targeting remain elusive. In this review, we evaluate previously-reported miRNA targeting rules and discuss the biological impact of the functional miRNAs on gene-regulatory networks as well as the future direction of miRNA targeting research. [BMB Reports 2017; 50(11): 554-559].

Korean Variant Archive (KOVA): a reference database of genetic variations in the Korean population.

Despite efforts to interrogate human genome variation through large-scale databases, systematic preference toward populations of Caucasian descendants has resulted in unintended reduction of power in studying non-Caucasians. Here we report a compilation of coding variants from 1,055 healthy Korean individuals (KOVA; Korean Variant Archive). The samples were sequenced to a mean depth of 75x, yielding 101 singleton variants per individual. Population genetics analysis demonstrates that the Korean population is a distinct ethnic group comparable to other discrete ethnic groups in Africa and Europe, providing a rationale for such independent genomic datasets. Indeed, KOVA conferred 22.8% increased variant filtering power in addition to Exome Aggregation Consortium (ExAC) when used on Korean exomes. Functional assessment of nonsynonymous variant supported the presence of purifying selection in Koreans. Analysis of copy number variants detected 5.2 deletions and 10.3 amplifications per individual with an increased fraction of novel variants among smaller and rarer copy number variable segments. We also report a list of germline variants that are associated with increased tumor susceptibility. This catalog can function as a critical addition to the pre-existing variant databases in pursuing genetic studies of Korean individuals.

General rules for functional microRNA targeting.

The functional rules for microRNA (miRNA) targeting remain controversial despite their biological importance because only a small fraction of distinct interactions, called site types, have been examined among an astronomical number of site types that can occur between miRNAs and their target mRNAs. To systematically discover functional site types and to evaluate the contradicting rules reported previously, we used large-scale transcriptome data and statistically examined whether each of approximately 2 billion site types is enriched in differentially downregulated mRNAs responding to overexpressed miRNAs. Accordingly, we identified seven non-canonical functional site types, most of which are novel, in addition to four canonical site types, while also removing numerous false positives reported by previous studies. Extensive experimental validation and significantly elevated 3' UTR sequence conservation indicate that these non-canonical site types may have biologically relevant roles. Our expanded catalog of functional site types suggests that the gene regulatory network controlled by miRNAs may be far more complex than currently understood.

The Brain-Enriched MicroRNA miR-9-3p Regulates Synaptic Plasticity and Memory.

UNLABELLED: MicroRNAs (miRNAs) are small, noncoding RNAs that posttranscriptionally regulate gene expression in many tissues. Although a number of brain-enriched miRNAs have been identified, only a few specific miRNAs have been revealed as critical regulators of synaptic plasticity, learning, and memory. miR-9-5p/3p are brain-enriched miRNAs known to regulate development and their changes have been implicated in several neurological disorders, yet their role in mature neurons in mice is largely unknown. Here, we report that inhibition of miR-9-3p, but not miR-9-5p, impaired hippocampal long-term potentiation (LTP) without affecting basal synaptic transmission. Moreover, inhibition of miR-9-3p in the hippocampus resulted in learning and memory deficits. Furthermore, miR-9-3p inhibition increased the expression of the LTP-related genes Dmd and SAP97, the expression levels of which are negatively correlated with LTP. These results suggest that miR-9-3p-mediated gene regulation plays important roles in synaptic plasticity and hippocampus-dependent memory. SIGNIFICANCE STATEMENT: Despite the abundant expression of the brain-specific microRNA miR-9-5p/3p in both proliferating and postmitotic neurons, most functional studies have focused on their role in neuronal development. Here, we examined the role of miR-9-5p/3p in adult brain and found that miR-9-3p, but not miR-9-5p, has a critical role in hippocampal synaptic plasticity and memory. Moreover, we identified in vivo binding targets of miR-9-3p that are involved in the regulation of long-term potentiation. Our study provides the very first evidence for the critical role of miR-9-3p in synaptic plasticity and memory in the adult mouse.

Temporal Landscape of MicroRNA-Mediated Host-Virus Crosstalk during Productive Human Cytomegalovirus Infection.

Temporal profiles of miRNA activity during productive virus infection can provide fundamental insights into host-virus interactions. Most reported miRNA targetome analyses in the context of virus infection have been performed in latently infected cells and lack reliable models for quantifying the suppression efficacy at specific miRNA target sites. Here, we identified highly competent temporal miRNA targetomes during lytic HCMV infection by using AGO-CLIP-seq together with a bioinformatic method that quantifies miRNA functionality at a specific target site, called ACE-scoring. The repression efficiency at target sites correlates with the magnitude of the ACE-score, and temporal HCMV-encoded miRNA targetomes identified by ACE-scoring were significantly enriched in functional categories involved in pathways central for HCMV biology. Furthermore, comparative analysis between human and viral miRNA targetomes supports the existence of intimate cooperation and co-targeting between them. Our holistic survey provides a valuable resource for understanding host-virus interactions during lytic HCMV infection.

Genomic Alterations in the RB Pathway Indicate Prognostic Outcomes of Early-Stage Lung Adenocarcinoma.

PURPOSE: To better understand the complete genomic architecture of lung adenocarcinoma. EXPERIMENTAL DESIGN: We used array experiments to determine copy number variations and sequenced the complete exomes of the 247 lung adenocarcinoma tumor samples along with matched normal cells obtained from the same patients. Fully annotated clinical data were also available, providing an unprecedented opportunity to assess the impact of genomic alterations on clinical outcomes. RESULTS: We discovered that genomic alternations in the RB pathway are associated with significantly shorter disease-free survival in early-stage lung adenocarcinoma patients. This association was also observed in our independent validation cohort. The current treatment guidelines for early-stage lung adenocarcinoma patients recommend follow-up without adjuvant therapy after complete resection, except for high-risk patients. However, our findings raise the interesting possibility that additional clinical interventions might provide medical benefits to early-stage lung adenocarcinoma patients with genomic alterations in the RB pathway. When examining the association between genomic mutation and histologic subtype, we uncovered the characteristic genomic signatures of various histologic subtypes. Notably, the solid and the micropapillary subtypes demonstrated great diversity in the mutated genes, while the mucinous subtype exhibited the most unique landscape. This suggests that a more tailored therapeutic approach should be used to treat patients with lung adenocarcinoma. CONCLUSIONS: Our analysis of the genomic and clinical data for 247 lung adenocarcinomas should help provide a more comprehensive genomic portrait of lung adenocarcinoma, define molecular signatures of lung adenocarcinoma subtypes, and lead to the discovery of useful prognostic markers that could be used in personalized treatments for early-stage lung adenocarcinoma patients.

mRNA destabilization is the dominant effect of mammalian microRNAs by the time substantial repression ensues.

MicroRNAs (miRNAs) regulate target mRNAs through a combination of translational repression and mRNA destabilization, with mRNA destabilization dominating at steady state in the few contexts examined globally. Here, we extend the global steady-state measurements to additional mammalian contexts and find that regardless of the miRNA, cell type, growth condition, or translational state, mRNA destabilization explains most (66%->90%) miRNA-mediated repression. We also determine the relative dynamics of translational repression and mRNA destabilization for endogenous mRNAs as a miRNA is induced. Although translational repression occurs rapidly, its effect is relatively weak, such that by the time consequential repression ensues, the effect of mRNA destabilization dominates. These results imply that consequential miRNA-mediated repression is largely irreversible and provide other insights into the nature of miRNA-mediated regulation. They also simplify future studies, dramatically extending the known contexts and time points for which monitoring mRNA changes captures most of the direct miRNA effects.