Human IRES Atlas: an integrative platform for studying IRES-driven translational regulation in humans

DMLS: an automated pipeline to extract the Drosophila modular transcription regulators and targets from massive literature articles

Transcription regulation in multicellular species is mediated by modular transcription factor (TF) binding site combinations termed cis-regulatory modules (CRMs). Such CRM-mediated transcription regulation determines the gene expression patterns during development. Biologists frequently investigate CRM transcription regulation on gene expressions. However, the knowledge of the target genes and regulatory TFs participating in the CRMs under study is mostly fragmentary throughout the literature. Researchers need to afford tremendous human resources to fully surf through the articles deposited in biomedical literature databases in order to obtain the information. Although several novel text-mining systems are now available for literature triaging, these tools do not specifically focus on CRM-related literature prescreening, failing to correctly extract the information of the CRM target genes and regulatory TFs from the literature. For this reason, we constructed a supportive auto-literature prescreener called Drosophila Modular transcription-regulation Literature Screener (DMLS) that achieves the following: (i) prescreens articles describing experiments on modular transcription regulation, (ii) identifies the described target genes and TFs of the CRMs under study for each modular transcription-regulation-describing article and (iii) features an automated and extendable pipeline to perform the task. We demonstrated that the final performance of DMLS in extracting the described target gene and regulatory TF lists of CRMs under study for given articles achieved test macro area under the ROC curve (auROC) = 89.7% and area under the precision-recall curve (auPRC) = 77.6%, outperforming the intuitive gene name-occurrence-counting method by at least 19.9% in auROC and 30.5% in auPRC. The web service and the command line versions of DMLS are available at https://cobis.bme.ncku.edu.tw/DMLS/  and  https://github.com/cobisLab/DMLS/, respectively. Database Tool URL: https://cobis.bme.ncku.edu.tw/DMLS/.

A novel function of the M2 muscarinic receptor

The M2 muscarinic receptor (M2R) is a prototypic class A G protein-coupled receptor (GPCR). Interestingly, Fasciani et al. recently identified an internal translation start site within the M2 receptor mRNA, directing the expression of a C-terminal receptor fragment. Elevated during cellular stress, this polypeptide localizes to mitochondria where it inhibits oxidative phosphorylation.

Predicting functional UTR variants by integrating region-specific features

The untranslated region (UTR) of messenger ribonucleic acid (mRNA), including the 5'UTR and 3'UTR, plays a critical role in regulating gene expression and translation. Variants within the UTR can lead to changes associated with human traits and diseases; however, computational prediction of UTR variant effect is challenging. Current noncoding variant prediction mainly focuses on the promoters and enhancers, neglecting the unique sequence of the UTR and thereby limiting their predictive accuracy. In this study, using consolidated datasets of UTR variants from disease databases and large-scale experimental data, we systematically analyzed more than 50 region-specific features of UTR, including functional elements, secondary structure, sequence composition and site conservation. Our analysis reveals that certain features, such as C/G-related sequence composition in 5'UTR and A/T-related sequence composition in 3'UTR, effectively differentiate between nonfunctional and functional variant sets, unveiling potential sequence determinants of functional UTR variants. Leveraging these insights, we developed two classification models to predict functional UTR variants using machine learning, achieving an area under the curve (AUC) value of 0.94 for 5'UTR and 0.85 for 3'UTR, outperforming all existing methods. Our models will be valuable for enhancing clinical interpretation of genetic variants, facilitating the prediction and management of disease risk.

RNA isoform expression landscape of the human dorsal root ganglion generated from long-read sequencing

ABSTRACT

Splicing is a posttranscriptional RNA processing mechanism that enhances genomic complexity by creating multiple isoforms from the same gene. We aimed to characterize the isoforms expressed in the human peripheral nervous system, with the goal of creating a resource to identify novel isoforms of functionally relevant genes associated with somatosensation and nociception. We used long-read sequencing to document isoform expression in the human dorsal root ganglia from 3 organ donors and validated in silico by confirming expression in short-read sequencing from 3 independent organ donors. Nineteen thousand five hundred forty-seven isoforms of protein-coding genes were detected and validated. We identified 763 isoforms with at least one previously undescribed splice junction. Previously unannotated isoforms of multiple pain-associated genes, including ASIC3, MRGPRX1, and HNRNPK, were identified. In the novel isoforms of ASIC3, a region comprising approximately 35% of the 5'UTR was excised. By contrast, a novel splice junction was used in isoforms of MRGPRX1 to include an additional exon upstream of the start codon, consequently adding a region to the 5'UTR. Novel isoforms of HNRNPK were identified, which used previously unannotated splice sites to both excise exon 14 and include a sequence in the 3' end of exon 13. This novel insertion is predicted to introduce a tyrosine phosphorylation site potentially phosphorylated by SRC. We also independently confirm a recently reported DRG-specific splicing event in WNK1 that gives insight into how painless peripheral neuropathy occurs when this gene is mutated. Our findings give a clear overview of mRNA isoform diversity in the human dorsal root ganglia obtained using long-read sequencing.

DEBFold: Computational Identification of RNA Secondary Structures for Sequences across Structural Families Using Deep Learning

It is now known that RNAs play more active roles in cellular pathways beyond simply serving as transcription templates. These biological mechanisms might be mediated by higher RNA stereo conformations, triggering the need to understand RNA secondary structures first. However, experimental protocols for solving RNA structures are unavailable for large-scale investigation due to their high costs and time-consuming nature. Various computational tools were thus developed to predict the RNA secondary structures from sequences. Recently, deep networks have been investigated to help predict RNA structures directly from their sequences. However, existing deep-learning-based tools are more or less suffering from model overfitting due to their complicated problem formulation and defective model training processes, limiting their applications across sequences from different structural families. In this research, we designed a two-stage RNA structure prediction strategy called DEBFold (deep ensemble boosting and folding) based on convolution encoding/decoding and self-attention mechanisms to enhance the existing thermodynamic structure models. Moreover, the model training process followed rigorous steps to achieve an acceptable prediction generalization. On the family-wise reserved test sets and the PDB-derived test set, DEBFold achieves better structure prediction performance over traditional tools and existing deep-learning methods. In summary, we obtained a cutting-edge deep-learning-based structure prediction tool with supreme across-family generalization performance. The DEBFold tool can be accessed at https://cobis.bme.ncku.edu.tw/DEBFold/.

Identifying Human miRNA Target Sites via Learning the Interaction Patterns between miRNA and mRNA Segments

miRNAs (microRNAs) target specific mRNA (messenger RNA) sites to regulate their translation expression. Although miRNA targeting can rely on seed region base pairing, animal miRNAs, including human miRNAs, typically cooperate with several cofactors, leading to various noncanonical pairing rules. Therefore, identifying the binding sites of animal miRNAs remains challenging. Because experiments for mapping miRNA targets are costly, computational methods are preferred for extracting potential miRNA-mRNA fragment binding pairs first. However, existing prediction tools can have significant false positives due to the prevalent noncanonical miRNA binding behaviors and the information-biased training negative sets that were used while constructing these tools. To overcome these obstacles, we first prepared an information-balanced miRNA binding pair ground-truth data set. A miRNA-mRNA interaction-aware model was then designed to help identify miRNA binding events. On the test set, our model (auROC = 94.4%) outperformed existing models by at least 2.8% in auROC. Furthermore, we showed that this model can suggest potential binding patterns for miRNA-mRNA sequence interacting pairs. Finally, we made the prepared data sets and the designed model available at http://cosbi2.ee.ncku.edu.tw/mirna_binding/download.

On the Inadequacy of the Current Transgenic Animal Models of Alzheimer's Disease: The Path Forward

For at least two reasons, the current transgenic animal models of Alzheimer's disease (AD) appear to be patently inadequate. They may be useful in many respects, the AD models; however, they are not. First, they are incapable of developing the full spectrum of the AD pathology. Second, they respond spectacularly well to drugs that are completely ineffective in the treatment of symptomatic AD. These observations indicate that both the transgenic animal models and the drugs faithfully reflect the theory that guided the design and development of both, the amyloid cascade hypothesis (ACH), and that both are inadequate because their underlying theory is. This conclusion necessitated the formulation of a new, all-encompassing theory of conventional AD-the ACH2.0. The two principal attributes of the ACH2.0 are the following. One, in conventional AD, the agent that causes the disease and drives its pathology is the intraneuronal amyloid-β (iAβ) produced in two distinctly different pathways. Two, following the commencement of AD, the bulk of Aβ is generated independently of Aβ protein precursor (AβPP) and is retained inside the neuron as iAβ. Within the framework of the ACH2.0, AβPP-derived iAβ accumulates physiologically in a lifelong process. It cannot reach levels required to support the progression of AD; it does, however, cause the disease. Indeed, conventional AD occurs if and when the levels of AβPP-derived iAβ cross the critical threshold, elicit the neuronal integrated stress response (ISR), and trigger the activation of the AβPP-independent iAβ generation pathway; the disease commences only when this pathway is operational. The iAβ produced in this pathway reaches levels sufficient to drive the AD pathology; it also propagates its own production and thus sustains the activity of the pathway and perpetuates its operation. The present study analyzes the reason underlying the evident inadequacy of the current transgenic animal models of AD. It concludes that they model, in fact, not Alzheimer's disease but rather the effects of the neuronal ISR sustained by AβPP-derived iAβ, that this is due to the lack of the operational AβPP-independent iAβ production pathway, and that this mechanism must be incorporated into any successful AD model faithfully emulating the disease. The study dissects the plausible molecular mechanisms of the AβPP-independent iAβ production and the pathways leading to their activation, and introduces the concept of conventional versus unconventional Alzheimer's disease. It also proposes the path forward, posits the principles of design of productive transgenic animal models of the disease, and describes the molecular details of their construction.

Combining a prioritization strategy and functional studies nominates 5'UTR variants underlying inherited retinal disease

BACKGROUND

5' untranslated regions (5'UTRs) are essential modulators of protein translation. Predicting the impact of 5'UTR variants is challenging and rarely performed in routine diagnostics. Here, we present a combined approach of a comprehensive prioritization strategy and functional assays to evaluate 5'UTR variation in two large cohorts of patients with inherited retinal diseases (IRDs).

METHODS

We performed an isoform-level re-analysis of retinal RNA-seq data to identify the protein-coding transcripts of 378 IRD genes with highest expression in retina. We evaluated the coverage of their 5'UTRs by different whole exome sequencing (WES) kits. The selected 5'UTRs were analyzed in whole genome sequencing (WGS) and WES data from IRD sub-cohorts from the 100,000 Genomes Project (n = 2397 WGS) and an in-house database (n = 1682 WES), respectively. Identified variants were annotated for 5'UTR-relevant features and classified into seven categories based on their predicted functional consequence. We developed a variant prioritization strategy by integrating population frequency, specific criteria for each category, and family and phenotypic data. A selection of candidate variants underwent functional validation using diverse approaches.

RESULTS

Isoform-level re-quantification of retinal gene expression revealed 76 IRD genes with a non-canonical retina-enriched isoform, of which 20 display a fully distinct 5'UTR compared to that of their canonical isoform. Depending on the probe design, 3-20% of IRD genes have 5'UTRs fully captured by WES. After analyzing these regions in both cohorts, we prioritized 11 (likely) pathogenic variants in 10 genes (ARL3, MERTK, NDP, NMNAT1, NPHP4, PAX6, PRPF31, PRPF4, RDH12, RD3), of which 7 were novel. Functional analyses further supported the pathogenicity of three variants. Mis-splicing was demonstrated for the PRPF31:c.-9+1G>T variant. The MERTK:c.-125G>A variant, overlapping a transcriptional start site, was shown to significantly reduce both luciferase mRNA levels and activity. The RDH12:c.-123C>T variant was found in cis with the hypomorphic RDH12:c.701G>A (p.Arg234His) variant in 11 patients. This 5'UTR variant, predicted to introduce an upstream open reading frame, was shown to result in reduced RDH12 protein but unaltered mRNA levels.

CONCLUSIONS

This study demonstrates the importance of 5'UTR variants implicated in IRDs and provides a systematic approach for 5'UTR annotation and validation that is applicable to other inherited diseases.

Exon Junction Complex Mediates the Cap-Independent Translation of Circular RNA

Evidence that circular RNAs (circRNA) serve as protein template is accumulating. However, how the cap-independent translation is controlled remains largely uncharacterized. Here, we show that the presence of intron and thus splicing promote cap-independent translation. By acquiring the exon junction complex (EJC) after splicing, the interaction between circRNA and ribosomes was promoted, thereby facilitating translation. Prevention of splicing by treatment with spliceosome inhibitor or mutating splicing signal hindered cap-independent translation of circRNA. Moreover, EJC-tethering using Cas13 technology reconstituted EJC-dependent circRNA translation. Finally, the level of a coding circRNA from succinate dehydrogenase assembly factor 2 (circSDHAF2) was found to be elevated in the tumorous tissues from patients with colorectal cancer, and shown to be critical in tumorigenesis of colorectal cancer in both cell and murine models. These findings reveal that EJC-dependent control of circSDHAF2 translation is involved in the regulation of oncogenic pathways.

IMPLICATIONS

EJC-mediated cap-independent translation of circRNA is implicated in the tumorigenesis of colorectal cancer.

RNA isoform expression landscape of the human dorsal root ganglion (DRG) generated from long read sequencing

Splicing is a post-transcriptional RNA processing mechanism that enhances genomic complexity by creating multiple isoforms from the same gene. Diversity in splicing in the mammalian nervous system is associated with neuronal development, synaptic function and plasticity, and is also associated with diseases of the nervous system ranging from neurodegeneration to chronic pain. We aimed to characterize the isoforms expressed in the human peripheral nervous system, with the goal of creating a resource to identify novel isoforms of functionally relevant genes associated with somatosensation and nociception. We used long read sequencing (LRS) to document isoform expression in the human dorsal root ganglia (hDRG) from 3 organ donors. Isoforms were validated in silico by confirming expression in hDRG short read sequencing (SRS) data from 3 independent organ donors. 19,547 isoforms of protein-coding genes were detected using LRS and validated with SRS and strict expression cutoffs. We identified 763 isoforms with at least one previously undescribed splice-junction. Previously unannotated isoforms of multiple pain-associated genes, including ASIC3, MRGPRX1 and HNRNPK were identified. In the novel isoforms of ASIC3, a region comprising ~35% of the 5'UTR was excised. In contrast, a novel splice-junction was utilized in isoforms of MRGPRX1 to include an additional exon upstream of the start-codon, consequently adding a region to the 5'UTR. Novel isoforms of HNRNPK were identified which utilized previously unannotated splice-sites to both excise exon 14 and include a sequence in the 5' end of exon 13. The insertion and deletion in the coding region was predicted to excise a serine-phosphorylation site favored by cdc2, and replace it with a tyrosine-phosphorylation site potentially phosphorylated by SRC. We also independently confirm a recently reported DRG-specific splicing event in WNK1 that gives insight into how painless peripheral neuropathy occurs when this gene is mutated. Our findings give a clear overview of mRNA isoform diversity in the hDRG obtained using LRS. Using this work as a foundation, an important next step will be to use LRS on hDRG tissues recovered from people with a history of chronic pain. This should enable identification of new drug targets and a better understanding of chronic pain that may involve aberrant splicing events.

RDDL: A systematic ensemble pipeline tool that streamlines balancing training schemes to reduce the effects of data imbalance in rare-disease-related deep-learning applications

Identifying lowly prevalent diseases, or rare diseases, in their early stages is key to disease treatment in the medical field. Deep learning techniques now provide promising tools for this purpose. Nevertheless, the low prevalence of rare diseases entangles the proper application of deep networks for disease identification due to the severe class-imbalance issue. In the past decades, some balancing methods have been studied to handle the data-imbalance issue. The bad news is that it is verified that none of these methods guarantees superior performance to others. This performance variation causes the need to formulate a systematic pipeline with a comprehensive software tool for enhancing deep-learning applications in rare disease identification. We reviewed the existing balancing schemes and summarized a systematic deep ensemble pipeline with a constructed tool called RDDL for handling the data imbalance issue. Through two real case studies, we showed that rare disease identification could be boosted with this systematic RDDL pipeline tool by lessening the data imbalance problem during model training. The RDDL pipeline tool is available at https://github.com/cobisLab/RDDL/.

The Amyloid Cascade Hypothesis 2.0 for Alzheimer's Disease and Aging-Associated Cognitive Decline: From Molecular Basis to Effective Therapy

With the long-standing amyloid cascade hypothesis (ACH) largely discredited, there is an acute need for a new all-encompassing interpretation of Alzheimer's disease (AD). Whereas such a recently proposed theory of AD is designated ACH2.0, its commonality with the ACH is limited to the recognition of the centrality of amyloid-β (Aβ) in the disease, necessitated by the observation that all AD-causing mutations affect, in one way or another, Aβ. Yet, even this narrow commonality is superficial since AD-causing Aβ of the ACH differs distinctly from that specified in the ACH2.0: Whereas in the former, the disease is caused by secreted extracellular Aβ, in the latter, it is triggered by Aβ-protein-precursor (AβPP)-derived intraneuronal Aβ (iAβ) and driven by iAβ generated independently of AβPP. The ACH2.0 envisions AD as a two-stage disorder. The first, asymptomatic stage is a decades-long accumulation of AβPP-derived iAβ, which occurs via internalization of secreted Aβ and through intracellular retention of a fraction of Aβ produced by AβPP proteolysis. When AβPP-derived iAβ reaches critical levels, it activates a self-perpetuating AβPP-independent production of iAβ that drives the second, devastating AD stage, a cascade that includes tau pathology and culminates in neuronal loss. The present study analyzes the dynamics of iAβ accumulation in health and disease and concludes that it is the prime factor driving both AD and aging-associated cognitive decline (AACD). It discusses mechanisms potentially involved in AβPP-independent generation of iAβ, provides mechanistic interpretations for all principal aspects of AD and AACD including the protective effect of the Icelandic AβPP mutation, the early onset of FAD and the sequential manifestation of AD pathology in defined regions of the affected brain, and explains why current mouse AD models are neither adequate nor suitable. It posits that while drugs affecting the accumulation of AβPP-derived iAβ can be effective only protectively for AD, the targeted degradation of iAβ is the best therapeutic strategy for both prevention and effective treatment of AD and AACD. It also proposes potential iAβ-degrading drugs.

Translation factor eIF5a is essential for IFNγ production and cell cycle regulation in primary CD8+ T lymphocytes

Control of mRNA translation adjusts protein production rapidly and facilitates local cellular responses to environmental conditions. Traditionally initiation of translation is considered to be a major translational control point, however, control of peptide elongation is also important. Here we show that the function of the elongation factor, eIF5a, is regulated dynamically in naïve CD8+ T cells upon activation by post-translational modification, whereupon it facilitates translation of specific subsets of proteins. eIF5a is essential for long-term survival of effector CD8+ T cells and sequencing of nascent polypeptides indicates that the production of proteins which regulate proliferation and key effector functions, particularly the production of IFNγ and less acutely TNF production and cytotoxicity, is dependent on the presence of functional eIF5a. Control of translation in multiple immune cell lineages is required to co-ordinate immune responses and these data illustrate that translational elongation contributes to post-transcriptional regulons important for the control of inflammation.

YMLA: A comparative platform to carry out functional enrichment analysis for multiple gene lists in yeast

Comparative analysis among multiple gene lists on their functional features is now a routine task due to the advancement of high-throughput experiments. Several enrichment analysis tools were developed in the past. However, these tools mainly focus on one gene list and contain only gene ontology or interaction features. What makes it worse, comparative investigation and customized feature set reanalysis are still unavailable. Therefore, we constructed the YMLA (Yeast Multiple List Analyzer) platform in this research. YMLA includes 39 yeast features and facilitates comparative analysis among multiple gene lists via tabular views, heatmaps, and network plots. Moreover, the customized feature set reanalysis function was implemented in YMLA to help form mechanism hypotheses based on a selected enriched feature subset. We demonstrated the biological applicability of YMLA via example lists consisting of genes with top/bottom translation efficiency values. The analysis results provided by YMLA reveal novel facts consistent with previous experiments. YMLA is available at https://cosbi7.ee.ncku.edu.tw/YMLA/.

Translation of SARS-CoV-2 gRNA Is Extremely Efficient and Competitive despite a High Degree of Secondary Structures and the Presence of an uORF

The SARS-CoV-2 infection generates up to nine different sub-genomic mRNAs (sgRNAs), in addition to the genomic RNA (gRNA). The 5'UTR of each viral mRNA shares the first 75 nucleotides (nt.) at their 5'end, called the leader, but differentiates by a variable sequence (0 to 190 nt. long) that follows the leader. As a result, each viral mRNA has its own specific 5'UTR in term of length, RNA structure, uORF and Kozak context; each one of these characteristics could affect mRNA expression. In this study, we have measured and compared translational efficiency of each of the ten viral transcripts. Our data show that most of them are very efficiently translated in all translational systems tested. Surprisingly, the gRNA 5'UTR, which is the longest and the most structured, was also the most efficient to initiate translation. This property is conserved in the 5'UTR of SARS-CoV-1 but not in MERS-CoV strain, mainly due to the regulation imposed by the uORF. Interestingly, the translation initiation mechanism on the SARS-CoV-2 gRNA 5'UTR requires the cap structure and the components of the eIF4F complex but showed no dependence in the presence of the poly(A) tail in vitro. Our data strongly suggest that translation initiation on SARS-CoV-2 mRNAs occurs via an unusual cap-dependent mechanism.

SSRTool: a web tool for evaluating RNA secondary structure predictions based on species-specific functional interpretability

RNA secondary structures can carry out essential cellular functions alone or interact with one another to form the hierarchical tertiary structures. Experimental structure identification approa ches can show the in vitro structures of RNA molecules. However, they usually have limits in the resolution and are costly. In silico structure prediction tools are thus primarily relied on for pre-experiment analysis. Various structure prediction models have been developed over the decades. Since these tools are usually used before knowing the actual RNA structures, evaluating and ranking the pile of secondary structure predictions of a given sequence is essential in computational analysis. In this research, we implemented a web service called SSRTool (RNA Secondary Structure prediction Ranking Tool) to assist in the ranking and evaluation of the generated predicted structures of a given sequence. Based on the computed species-specific interpretability significance in four common RNA structure–function aspects, SSRTool provides three functions along with visualization interfaces: (1) Rank user-generated predictions. (2) Provide an automated streamline of structure prediction and ranking for a given sequence. (3) Infer the functional aspects of a given structure. We demonstrated the applicability of SSRTool via real case studies and reported the similar trends between computed species-specific rankings and the corresponding prediction F1 values. The SSRTool web service is available online at https://cobisHSS0.im.nuk.edu.tw/SSRTool/, http://cosbi3.ee.ncku.edu.tw/SSRTool/, or the redirecting site https://github.com/cobisLab/SSRTool/.

KDmarkers: A biomarker database for investigating epigenetic methylation and gene expression levels in Kawasaki disease

Kawasaki disease (KD) is a form of acute systemic vasculitis that primarily affects children and has become the most common cause of acquired heart disease. While the etiopathogenesis of KD remains unknown, the diagnostic criteria of KD have been well established. Nevertheless, the diagnosis of KD is currently based on subjective clinical symptoms, and no molecular biomarker is yet available. We have previously performed and combined methylation array (Illumina HumanMethylation450 BeadChip) and transcriptome array (Affymetrix GeneChip Human Transcriptome Array 2.0) to identify genes that are differentially methylated/expressed in KD patients compared with control subjects. We have found that decreased methylation levels combined with elevated gene expression can indicate genes (e.g., toll-like receptors and CD177) involved in the disease mechanisms of KD. In this study, we constructed a database called KDmarkers to allow researchers to access these valuable potential KD biomarkers identified via methylation array and transcriptome array. KDmarkers provides three search modes. First, users can search genes differentially methylated and/or differentially expressed in KD patients compared with control subjects. Second, users can check the KD patient groups in which a given gene is differentially methylated and/or differentially expressed. Third, users can explore the DNA methylation levels and gene expression levels in all samples (KD patients and controls) for a particular gene of interest. We further demonstrated that the results in KDmarkers are strongly associated with KD immune responses. All analysis results can be downloaded for downstream experimental designs. KDmarkers is available online at https://cosbi.ee.ncku.edu.tw/KDmarkers/.

Reversal of G-Quadruplexes’ Role in Translation Control When Present in the Context of an IRES

G-quadruplexes (GQs) are secondary nucleic acid structures that play regulatory roles in various cellular processes. G-quadruplex-forming sequences present within the 5′ UTR of mRNAs can function not only as repressors of translation but also as elements required for optimum function. Based upon previous reports, the majority of the 5′ UTR GQ structures inhibit translation, presumably by blocking the ribosome scanning process that is essential for detection of the initiation codon. However, there are certain mRNAs containing GQs that have been identified as positive regulators of translation, as they are needed for translation initiation. While most cellular mRNAs utilize the 5′ cap structure to undergo cap-dependent translation initiation, many rely on cap-independent translation under certain conditions in which the cap-dependent initiation mechanism is not viable or slowed down, for example, during development, under stress and in many diseases. Cap-independent translation mainly occurs via Internal Ribosomal Entry Sites (IRESs) that are located in the 5′ UTR of mRNAs and are equipped with structural features that can recruit the ribosome or other factors to initiate translation without the need for a 5′ cap. In this review, we will focus only on the role of RNA GQs present in the 5′ UTR of mRNAs, where they play a critical role in translation initiation, and discuss the potential mechanism of this phenomenon, which is yet to be fully delineated.

RNA-Binding Proteins as Regulators of Internal Initiation of Viral mRNA Translation

Viruses are obligate intracellular parasites that depend on the host’s protein synthesis machinery for translating their mRNAs. The viral mRNA (vRNA) competes with the host mRNA to recruit the translational machinery, including ribosomes, tRNAs, and the limited eukaryotic translation initiation factor (eIFs) pool. Many viruses utilize non-canonical strategies such as targeting host eIFs and RNA elements known as internal ribosome entry sites (IRESs) to reprogram cellular gene expression, ensuring preferential translation of vRNAs. In this review, we discuss vRNA IRES-mediated translation initiation, highlighting the role of RNA-binding proteins (RBPs), other than the canonical translation initiation factors, in regulating their activity.

regCNN: identifying Drosophila genome-wide cis-regulatory modules via integrating the local patterns in epigenetic marks and transcription factor binding motifs

Transcription regulation in metazoa is controlled by the binding events of transcription factors (TFs) or regulatory proteins on specific modular DNA regulatory sequences called cis-regulatory modules (CRMs). Understanding the distributions of CRMs on a genomic scale is essential for constructing the metazoan transcriptional regulatory networks that help diagnose genetic disorders. While traditional reporter-assay CRM identification approaches can provide an in-depth understanding of functions of some CRM, these methods are usually cost-inefficient and low-throughput. It is generally believed that by integrating diverse genomic data, reliable CRM predictions can be made. Hence, researchers often first resort to computational algorithms for genome-wide CRM screening before specific experiments. However, current existing in silico methods for searching potential CRMs were restricted by low sensitivity, poor prediction accuracy, or high computation time from TFBS composition combinatorial complexity. To overcome these obstacles, we designed a novel CRM identification pipeline called regCNN by considering the base-by-base local patterns in TF binding motifs and epigenetic profiles. On the test set, regCNN shows an accuracy/auROC of 84.5%/92.5% in CRM identification. And by further considering local patterns in epigenetic profiles and TF binding motifs, it can accomplish 4.7% (92.5%–87.8%) improvement in the auROC value over the average value-based pure multi-layer perceptron model. We also demonstrated that regCNN outperforms all currently available tools by at least 11.3% in auROC values. Finally, regCNN is verified to be robust against its resizing window hyperparameter in dealing with the variable lengths of CRMs. The model of regCNN can be downloaded athttp://cobisHSS0.im.nuk.edu.tw/regCNN/.

YTLR: Extracting yeast transcription factor-gene associations from the literature using automated literature readers

Cells adapt to environmental stresses mainly via transcription reprogramming. Correct transcription control is mediated by the interactions between transcription factors (TF) and their target genes. These TF-gene associations can be probed by chromatin immunoprecipitation techniques and knockout experiments, revealing TF binding (TFB) and regulatory (TFR) evidence, respectively. Nevertheless, most evidence is still fragmentary in the literature and requires tremendous human resources to curate. We developed the first pipeline called YTLR (Yeast Transcription-regulation Literature Reader) to automate TF-gene relation extraction from the literature. YTLR first identifies articles with TFB and TFR information. Then TF-gene binding pairs are extracted from the TFB articles, and TF-gene regulatory associations are recognized from the TFR papers. On gathered test sets, YTLR achieves an AUC value of 98.8% in identifying articles with TFB evidence and AUC = 83.4% in extracting the detailed TF-gene binding pairs. And similarly, YTLR also obtains an AUC value of 98.2% in identifying TFR articles and AUC = 80.4% in extracting the detailed TF-gene regulatory associations. Furthermore, YTLR outperforms previous methods in both tasks. To facilitate researchers in extracting TF-gene transcriptional relations from large-scale queried articles, an automated and easy-to-use software tool based on the YTLR pipeline is constructed. In summary, YTLR aims to provide easier literature pre-screening for curators and help researchers gather yeast TF-gene transcriptional relation conclusions from articles in a high-throughput fashion. The YTLR pipeline software tool can be downloaded at https://github.com/cobisLab/YTLR/.

Identifying piRNA targets on mRNAs in C. elegans using a deep multi-head attention network

Background

Piwi-interacting RNAs (piRNAs) are the small non-coding RNAs (ncRNAs) that silence genomic transposable elements. And researchers found out that piRNA also regulates various endogenous transcripts. However, there is no systematic understanding of the piRNA binding patterns and how piRNA targets genes. While various prediction methods have been developed for other similar ncRNAs (e.g., miRNAs), piRNA holds distinctive characteristics and requires its own computational model for binding target prediction.

Results

Recently, transcriptome-wide piRNA binding events in C. elegans were probed by PRG-1 CLASH experiments. Based on the probed piRNA-messenger RNAs (mRNAs) binding pairs, in this research, we devised the first deep learning architecture based on multi-head attention to computationally identify piRNA targeting mRNA sites. In the devised deep network, the given piRNA and mRNA segment sequences are first one-hot encoded and undergo a combined operation of convolution and squeezing-extraction to unravel motif patterns. And we incorporate a novel multi-head attention sub-network to extract the hidden piRNA binding rules that can simulate the biological piRNA target recognition process. Finally, the true piRNA–mRNA binding pairs are identified by a deep fully connected sub-network. Our model obtains a supreme discriminatory power of AUC \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$=$$\end{document}= 93.3% on an independent test set and successfully extracts the verified binding pattern of a synthetic piRNA. These results demonstrated that the devised model achieves high prediction performance and suggests testable potential biological piRNA binding rules.

Conclusions

In this research, we developed the first deep learning method to identify piRNA targeting sites on C. elegans mRNAs. And the developed deep learning method is demonstrated to be of high accuracy and can provide biological insights into piRNA–mRNA binding patterns. The piRNA binding target identification network can be downloaded from http://cosbi2.ee.ncku.edu.tw/data_download/piRNA_mRNA_binding.

Cancer DEIso: An integrative analysis platform for investigating differentially expressed gene-level and isoform-level human cancer markers

Transcript isoforms regulated by alternative splicing can substantially impact carcinogenesis, leading to a need to obtain clues for both gene differential expression and malfunctions of isoform distributions in cancer studies. The Cancer Genome Atlas (TCGA) project was launched in 2008 to collect cancer-related genome mutation raw data from the population. While many repositories tried to add insights into the raw data in TCGA, no existing database provides both comprehensive gene-level and isoform-level cancer stage marker investigation and survival analysis. We constructed Cancer DEIso to facilitate in-depth analyses for both gene-level and isoform-level human cancer studies. Patient RNA-seq data, sample sheets, patient clinical data, and human genome datasets were collected and processed in Cancer DEIso. And four functions to search differentially expressed genes/isoforms between cancer stages were implemented: (i) Search potential gene/isoform markers for a specified cancer type and its two stages; (ii) Search potentially induced cancer types and stages for a gene/isoform; (iii) Expression survival analysis on a given gene/isoform for some cancer; (iv) Gene/isoform stage expression comparison visualization. As an example, we demonstrate that Cancer DEIso can indicate potential colorectal cancer isoform diagnostic markers that are not easily detected when only gene-level expressions are considered. Cancer DEIso is available at http://cosbi4.ee.ncku.edu.tw/DEIso/.