RNA

Nearest neighbor parameters are widely used in software for estimating the conformational stability of an RNA sequence folding into a specific structure. Folding stability for RNA with canonical nucleotides A, C, G, and U has been widely studied, but the same is not true for most modified nucleotides. In this work, we present a comprehensive set of nearest neighbor parameters for estimating the folding stability of RNA including pseudouridine in helical or loop contexts. These parameters are derived from 210 optical melting experiments involving helices with pseudouridine-A and pseudouridine-G pairs and with pseudouridine in loop motifs. The experiments include sequences with pseudouridine and U in the same strand, including U-A and U-G pairs, allowing us to consider the folding stability of sequences with both U and pseudouridine. On average, pseudouridine stabilizes RNA folding compared to U in an analogous motif, although this effect is sequence-context dependent. These parameters improve the modeling of folding stability for RNA secondary structures containing pseudouridine. We demonstrate that these parameters successfully model the secondary structure change for Saccharomyces cerevisiae U2 snRNA when two additional inducible pseudouridines are present. These parameters are freely available and incorporated into the RNAstructure software package. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=81 SRC="FIGDIR/small/725682v1_ufig1.gif" ALT="Figure 1"> View larger version (14K): org.highwire.dtl.DTLVardef@e1167aorg.highwire.dtl.DTLVardef@18ac7f0org.highwire.dtl.DTLVardef@4c909eorg.highwire.dtl.DTLVardef@aa8bca_HPS_FORMAT_FIGEXP M_FIG C_FIG

Splicing is governed by interactions between the spliceosome and precursor RNA sequence and structural elements. However, the relative contributions of RNA sequence and structural elements remain unclear. Here, we systematically dissect these determinants using a high-throughput mutagenesis approach with the MAP3K7 intron reporter. The MAP3K7 gene encodes a serine/threonine kinase involved in response to environmental stress. MAP3K7 precursor RNA contains a cryptic 3 splice site that increases in use when the core spliceosomal protein SF3B1 is mutated. SF3B1 mutations are known to promote aberrant splicing and are associated with cancer, particularly the lysine 700 to glutamate mutation (K700E). We designed a pooled library of 249 MAP3K7 mutants targeting branch points, RNA-binding protein motifs, nucleotide composition and predicted structural elements. The impact of these mutants on splicing was measured in the context of normal and SF3B1 K700E expression. RNA structure was assessed in parallel using in vitro high-throughput SHAPE-MAP chemical probing. We found that branchpoint mutations drive the strongest increases in cryptic splice-site use. There is no overall correlation between cryptic splice-site use and structural similarity to the wild-type MAP3K7 RNA. However, mutants within an RNA binding protein hotspot (containing U2AF2, U2AF1, KHSRP and SRSF2 sites) are associated with cryptic splice-site use and structural similarity to wild-type MAP3K7 RNA. These structural changes are associated with increased ensemble diversity. Our results demonstrate that although there are key structured regions within an RNA, there is also extensive variability where divergent RNA structures allow for accurate splicing.

Codon optimality promotes efficient translation and, as recent research has shown, also extends mRNA lifetimes. However, how control is distributed between translation and mRNA degradation remains unclear. We show that this relative impact depends strongly on the measurement approach. Using fluorescent protein reporters can underestimate codon-optimality-dependent increases in translation efficiency. Conversely, analyses based on poly(A)-selected RNA overestimate the impact on translation, because stable transcripts undergoing poly(A) shortening are often inefficiently captured, leading to skewed protein-to-mRNA ratios. This technical bias is not offset by the marginal decline in ribosomal association observed as mRNAs age. Estimates based on total RNA measurements redistribute some of the control attributed to translation to mRNA stability, making the contributions comparable for mRNAs with shorter coding sequences. For longer mRNAs, codon optimality increasingly controls elongation speed, with a greater effect on translation efficiency than on degradation. These insights highlight the importance of measurement strategy for accurately quantifying the determinants of mRNA stability and protein synthesis.

Mutations in the low complexity domains of RNA-binding proteins are associated with neurodegenerative disease pathology. The TIA1 RNA-binding protein harbors seven such mutations linked to a clinical cohort of ALS patients. The altered low complexity domain sequence increases the number of TIA1-rich stress granules in cultured cells, delays their disassembly, and is associated with increased fibril formation. Altered molecular motions and contacts in condensed states like stress granules that result in the formation of amyloid-like fibril states is commonly observed for RNA-binding biomolecular condensates. Here we focus on the influence of the ALS mutations on fibril formation of the TIA1 low complexity domain. Repetitive seeding preparations of the seven TIA1 protein mutants all yield amyloid-like fibrils based on transmission electron microscopy images and increased thioflavin T fluorescence. Analysis of solid state nuclear magnetic resonance spectra recorded on all seven mutant fibrils reveals distinct structural differences in the relative to wild-type fibrils. Our results shed light on how the mutations affect structural conformations accessible to the TIA1 low complexity domain.

Polycomb Repressive Complex 2 (PRC2) maintains repression of genes specific for other cell differentiation states. PRC2 binds RNA in vitro with a preference for G-rich sequences. UV-based crosslinking coupled with immunoprecipitation (CLIP) experiments have shown that PRC2 also binds RNA in cells. Recently, Guo et al reported that a stringent denaturing variant of CLIP called CLAP did not detect PRC2 RNA binding in cells. We present a reanalysis of CLAP data that supports direct interaction of PRC2 with RNA in cells. CLAP for Halo-tagged PRC2 subunits from mixed populations of human and mouse cells specifically enriched for RNA from the species in which the proteins were tagged. The lack of apparent PRC2 RNA binding in Guo and colleagues analysis stems from a scaling step that deflates enrichment scores for low-complexity CLAP samples. Our findings pave the way for studies seeking to determine the physiological roles of PRC2 RNA binding activity.

The RNA demethylase FTO erases N6-methyladenosine (m6A) and cap-associated N6,2'-O-dimethyladenosine (m6Am) modifications. However, the molecular basis of its substrate selectivity and the biological effects of m6A versus m6Am demethylation in cells remain poorly understood. Here we report two engineered FTO separation-of-function mutants to selectively demethylate either m6A or m6Am modifications on RNA. While investigating the propensity of FTO active site residues to undergo self-hydroxylation, we found that mutations of FTO residue L203 resulted in impaired m6A demethylation but retained wild-type levels of m6Am demethylation, and that FTO L203A could function as a selective m6Am demethylase. Conversely, building on our recent work that identified conserved aromatic residues on FTO involved in mRNA 5' cap recognition, we found that the FTO H232A/W278A double mutant efficiently demethylates m6A modifications while exhibiting substantially impaired m6Am demethylation, making it a selective m6A demethylase. Together, these complementary FTO variants represent the first set of engineered mutations that shift FTO demethylation selectivity between m6A and m6Am substrates. These tools enable selective enzymatic removal of m6A or m6Am modifications in vitro for sequencing applications, and may facilitate understanding of FTO-mediated m6A versus m6Am demethylation in cellular and disease model systems.

We previously found that high genome replication fitness of the hepatitis C virus (HCV) was associated with severe disease in immunocompromised patients. Elevated replication fitness was mediated by accumulation of mutations in the replication enhancing domain (ReED) within domain (D) 2 of non-structural protein (NS) 5A. NS5A is a partially unstructured phosphoprotein lacking enzymatic activity but fulfilling a key role in HCV replication due to interacting with various cellular and viral proteins. It can exist in a variety of dimeric and oligomeric conformations mediated by NS5A D1 with clinically approved NS5A inhibitors proposed to exert their antiviral function by fixing these dimers in distinct conformations. In this study, we aimed at elucidating the ReEDs mode of action. AlphaFold modelling indicated a so far unrecognized NS5A dimerization site in the ReED. Indeed, split nano luciferase assays revealed a significantly stronger NS5A dimerization of high replicator ReED variants, suggesting that high replication fitness is mediated by enforcement of NS5A self-interaction. This hypothesis was supported by the effect of low dose (1 pM) NS5A inhibitor treatment, increasing replication fitness and phenocopying the effects of ReED mutations. Furthermore, we found that HCV isolate JFH1, replicating with very high efficiency, is completely resistant to the regulatory function of the ReED. Chimeric replicons composed of ReED resistant JFH1 and the ReED sensitive isolate J6 identified NS3 helicase and NS5B polymerase as critical genetic elements mediating ReED sensitivity/resistance. Our data overall suggest that NS5A is a negative regulator of HCV replication fitness with dimerization releasing the inhibitory interaction with helicase and/or polymerase, thereby likely facilitating initiation of RNA synthesis.

Ribosome biogenesis is a highly coordinated pathway that involves the assembly of ribosomal RNAs (rRNAs) with ribosomal proteins (r-proteins) to generate functional ribosomal subunits (r-subunits). The Saccharomyces cerevisiae (yeast) large 60S r-subunit consists of three rRNA molecules and 46 r-proteins. The contributions of nearly all r-proteins of the yeast large r-subunit have been characterized; however, a few non-essential proteins remain poorly understood. Although non-essential, human eL22 has been identified as a key player in p53 regulation during ribosomal stress and as a highly mutated target in cancers. Despite this function, the role of eL22 in ribosome maturation is still ill-defined. In this study, we characterized yeast eL22 r-protein. Our results show that eL22 assembles into intermediate nucleolar pre-60S ribosomal particles. Loss of eL22 impairs cell growth and reduces 60S r-subunit accumulation, phenotypes that are exacerbated at low temperatures. Analysis of pre-rRNA processing by pulse-chase labeling, northern blot hybridization, and primer extension reveals a defect in 27S pre-rRNA maturation, specifically at the level of 27SB pre-rRNA processing. Consequently, nuclear export of eL22-deficient pre-60S particles is mildly impaired. Furthermore, we identify genetic interactions between eL22 and neighboring r-proteins, eL38 and eL31. We conclude that eL22 assembly is required for optimal pre-60S maturation during middle nucleolar stages, particularly at low temperatures, a function likely supported by the cooperative action of other r-proteins associated with common elements of 25S rRNA. HighlightsO_LIWe have studied the role of r-protein eL22 in yeast ribosome assembly. C_LIO_LIeL22 is required for 60S ribosomal subunit production. C_LIO_LIThe absence of eL22 is critical at low temperatures. C_LIO_LIeL22 is important for 27SB pre-rRNA processing and nuclear export of pre-ribosomes. C_LIO_LIeL22 functionally interacts with r-proteins eL38 and eL31 in domain III of 25S rRNA. C_LI

Double-stranded RNA (dsRNA) is recognized by cellular receptors as a sign of viral infection, triggering the innate immune response. Increasing evidence shows that cellular dysregulation, for example in immune disorders and neurodegenerative diseases, can also lead to accumulation of endogenously produced dsRNA that stimulates a viral-like immune response. Additionally, dsRNA contamination in RNA therapeutics can lead to harmful side effects via a similar pathway. Despite the clinical relevance of dsRNA, reliable tools for its detection remain limited. At present, dsRNA detection relies almost exclusively on the monoclonal antibodies J2 and K1, which suffer from sequence bias and low sensitivity, limiting their reliability. To address this challenge, we aimed to repurpose naturally occurring dsRNA-binding domains (dsRBDs) to produce reliable, pan-specific affinity reagents for dsRNA. We first systematically screened the dsRBDs of the three human adenosine deaminases acting on RNA (ADARs). This analysis identified ADAR3 dsRBDs as promising candidates due to their reduced sequence dependence compared to the dsRBDs of ADAR1 and ADAR2. We then engineered ADAR3-derived dsRBD constructs having varying linker lengths and domain combinations, allowing us to specifically vary the length cutoff of dsRNA detected, thus creating dsRNA accumulation detected by ADAR3 RBDs (dsRADAR) affinity reagents. Finally, we demonstrate the superior performance of dsRADAR over currently available dsRNA antibodies in a cell model of viral infection and a tissue model of gastric inflammation. Together, dsRADAR provides a sensitive and reliable approach for imaging and quantifying diverse dsRNA structures in a variety of biological contexts. Graphic Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=124 SRC="FIGDIR/small/724404v1_ufig1.gif" ALT="Figure 1"> View larger version (24K): org.highwire.dtl.DTLVardef@1d89c30org.highwire.dtl.DTLVardef@1f64fc1org.highwire.dtl.DTLVardef@1ee391forg.highwire.dtl.DTLVardef@e834a6_HPS_FORMAT_FIGEXP M_FIG C_FIG

RNA helicases encoded by positive-strand RNA viruses are essential for genome replication, yet the specific biological functions and mechanochemical basis underlying these functions remain poorly defined. Progress has been limited by the difficulty of resolving individual catalytic steps under single-turnover conditions, which are often experimentally inaccessible for viral enzymes. Alphaviruses replicate within membrane-bound spherules that may alter local metabolite concentrations, raising the possibility that the enzymatic properties of alphaviral proteins differ from those of viruses with greater cytosolic exposure. Here, we present a kinetic and binding analysis of full-length non-structural protein 2 (nsP2) from Chikungunya virus, a multifunctional superfamily 1B NTPase and RNA helicase. Purified nsP2 binds nucleoside triphosphates with high affinity, exhibiting equilibrium dissociation constants in the single digit micromolar range. This property enabled single-turnover, pre-steady-state, and isotope-trapping experiments that are rarely feasible for viral helicases. These analyses identified two sequential conformational-change steps required for nucleotide hydrolysis. Molecular dynamics simulations suggest tightening of the RecA1 and RecA2 domains upon ATP binding followed by compaction of the enzyme mediated by interactions between the 1B subdomain and RecA2 domain. Product inhibition patterns support random release of ADP and inorganic phosphate, with relative binding affinities indicating that ADP dissociates first. The reaction is irreversible. Although nsP2 binds RNA tightly, strand separation under single-turnover conditions is too slow to represent ATP-driven unwinding, instead likely reflecting formation of an unwinding-competent nsP2-RNA complex. Together, these findings establish a quantitative framework for nsP2 function and provide a roadmap for mechanistic studies of alphaviral helicases. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=63 SRC="FIGDIR/small/723793v1_ufig1.gif" ALT="Figure 1"> View larger version (18K): org.highwire.dtl.DTLVardef@13899a1org.highwire.dtl.DTLVardef@ee1aadorg.highwire.dtl.DTLVardef@1991e1org.highwire.dtl.DTLVardef@b877f6_HPS_FORMAT_FIGEXP M_FIG C_FIG

Rfam is a comprehensive database of non-coding RNA (ncRNA) families providing curated sequence alignments, consensus secondary structures, and covariance models for thousands of RNA families. The database is essential for identifying structured non-coding RNAs in newly sequenced genomes and understanding RNA structure-function relationships. Here we present computational protocols for automated ncRNA annotation of viral genomes, and for programmatic interaction with Rfam through its RESTful API. We showcase genome-wide RNA structure visualization from a genome sequence and from a multiple sequence alignment by generating comprehensive 2D structure diagrams using newly developed features in R2DT. We also present practical examples for retrieving family metadata, downloading alignments, accessing secondary structures, and searching user sequences from the Rfam API. These methods enable researchers in virology and RNA biology to integrate Rfam data into custom bioinformatics pipelines, comparative analyses, and machine learning workflows.

Pseudouridine ({Psi}) is an important post-transcriptional modification of many noncoding RNAs that is under-characterized in microRNA (miRNA) due to historical limitations in pseudouridine mapping methods. {Psi} modification stabilizes RNA duplex structures and could therefore play an important role in miRNA target binding and repression. To investigate the extent to which mammalian miRNAs are modified with {Psi}, we profiled the modification landscape of short (<30 nt) RNA in human cells and mouse tissues using bisulfite sequencing. Our approach was powered to detect small RNA pseudouridylation based on robust detection of known {Psi} positions in tRNA fragments (tRFs), some of which show tissue-specific patterns of modification. In contrast with tRFs, we find that miRNA pseudouridylation is exceedingly rare, with a single modified miRNA (miR-3068-5p) identified in mouse tissues. Pseudouridylated miR-3068-5p diSerentially repressed predicted miRNA targets with less stable miRNA:mRNA pairing modes. This study fills a long-standing gap in transcriptome-wide {Psi} profiling and reveals a new potential function for {Psi} as a modulator of activity of small regulatory RNAs.

Variants affecting RNA splicing are a major contributor to human disease, yet the consequences of variants outside of the canonical splice motifs are often difficult to determine. Here, we present a protocol for minigene-based evaluation of candidate splice-altering variants. The methodology described includes locus-specific insert design, commercial gene fragment synthesis, and long-read sequencing. The combined approach enables rapid assay development and nucleotide level resolution of the effect on splice isoforms in vitro, providing a scalable framework for functional validation of predicted cryptic splice variants. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=197 SRC="FIGDIR/small/723105v1_ufig1.gif" ALT="Figure 1"> View larger version (42K): org.highwire.dtl.DTLVardef@1a88cb5org.highwire.dtl.DTLVardef@adda98org.highwire.dtl.DTLVardef@1ea587corg.highwire.dtl.DTLVardef@574a63_HPS_FORMAT_FIGEXP M_FIG C_FIG

Eukaryotes use several distinct quality control pathways to resolve aberrant ribosomes and mRNAs. For example, the no-go decay mRNA pathway is stimulated after ribosome collisions caused by stalled ribosomes translating damaged or truncated mRNAs. Separate decay pathways for non-functional 40S and 60S subunits containing rRNA mutations affecting decoding and peptidyl transferase activity, respectively, have also been elucidated. To our knowledge, whether eukaryotes have evolved a quality control pathway to sense and process globally stalled ribosomes is unclear; however, such a pathway would be advantageous to eukaryotes during exposure to natural elongation inhibitors such as ricin and diphtheria toxin. Here, we test how prolonged robust inhibition of elongation using a high dose of cycloheximide (CHX) affects ribosome turnover. Despite no decrease in cell viability and that mammalian ribosomes have been classically characterized of having a half-life of 3-5 days, a single 24 hr high dose of CHX resulted in drastically shortened half-lives (<24 hr) of 28S and 18S rRNA in A549 cells. A [~]2-fold reduction in nearly all ribosome species was observed by polysome analysis in HeLa and A549 cells after prolonged CHX treatment. Depletion of ribosomes was also evident when assessing ribosomal proteins from both the 40S and 60S subunits by Western blot. Literature supports that ribosomes can be degraded by autophagy and the ubiquitin (Ub)-proteasome system. Upon testing inhibitors of both pathways, only proteasome inhibitors (i.e., MG132 and bortezomib) rescued both rRNA and ribosomal protein levels. Proteasome inhibitors also rescued ribosome levels in polysome profiling experiments. Remarkably, rRNA levels were not rescued during CHX treatment when co-treated with the Ub activating enzyme E1 inhibitor, TAK243. Polysome analysis also showed that the high prolonged dose of CHX did not cause robust accumulation of collided ribosomes compared to control treatments. Proteasome-dependent turnover of rRNA was also observed with high doses of other elongation inhibitors, namely anisomycin, homoharringtonine, and lactimidomycin. The recognition capabilities of the pathway were further expanded as we observed that 80S ribosomes not trapped on the mRNA were also targeted for degradation by the proteasome. Together, our findings define the framework of a regulatory pathway in mammalian cells that degrades both ribosomal subunits in response to prolonged periods of robust elongation inhibition.

25S nonfunctional RNA decay (NRD) eliminates 60S ribosomal subunits carrying inactivating mutations in the RNA. However, how cells identify defective subunits has not been described. We recently showed that the zinc-finger protein Reh1 is the last assembly factor to be released from a nascent 60S subunit. We now show that in yeast Reh1 is required for the degradation of 25S NRD substrates. 25S rRNAs carrying mutations in the catalytic center, A2820G or U2954A (A2451 and U2585, respectively in E coli numbering), are unstable in wildtype cells but are fully stabilized when REH1 is deleted. However, not all 25S rRNA mutations are recognized by Reh1. Ribosomes with a truncated L1 stalk engage in translation but cannot support viability. These ribosomes display a half-life indistinguishable from wild-type rRNA, suggesting that yeast does not have a robust surveillance system for such mutant ribosomes. Deletion of REH1 also has no impact on the levels of defective 18S rRNA. These results indicate that Reh1 and 25S NRD are specific for mutations in or near the catalytic center of the ribosome.

The long non-coding RNA MALAT1 is a conserved oncogenic driver whose function relies on a 3 triple-helix motif. While its biochemistry is well-characterized in vitro, the endogenous requirement for this motif in regulating the stability of the transcript and other genes residing in its locus remains unclear. In this study, we employed a dual-sgRNA CRISPR-Cas9 approach to systematically excise triple-helix-forming sequences from the native MALAT1 locus in gastric (AGS) and breast (MCF7) cancer cells. Our findings demonstrate that the 3 end functions as a binary structural switch. Any perturbation (ranging from genomic deletions to a single-base insertion) triggers total transcript collapse and rapid exonucleolytic decay. This instability leads to locus-wide transcriptomic failure, characterized by the precipitous loss of the antisense transcript TALAM1, while the biogenesis of the small RNA mascRNA (a byproduct of MALAT1, also involved in cancer) remains decoupled and unaffected. In cellulo, DMS probing reveals that edited transcripts retain structural complexity. Phenotypically, structural disruption of the 3 end significantly impairs the proliferative capacity of both cancer cellular models. These results identify the 3 triple-helix as an indispensable determinant of MALAT1 stability and provide endogenous validation for its role in cancer cells.

Ribozyme-based permuted intron-exon (PIE) systems offer a protein-independent route to circRNA production, but existing platforms require elevated temperatures that promote RNA degradation. Here we report the first application of the Candida albicans mitochondrial large subunit (C.a.mtLSU) group I intron as a PIE platform for circRNA synthesis, which we term PCanPIE (Pyle lab Candida PIE). We evaluated three peripheral stems, P5, P6b, and P8, as permutation sites and demonstrated that all three support circularization under near-physiological conditions (25{degrees}C, 6 mM MgCl2), without the 55{degrees}C heating step required by existing PIE systems. Kinetic analysis revealed that permutation site does not affect the observed splicing rate constant but does influence PCanPIE folding and therefore influences circularization efficiency. The P6b permutation yielded the highest circularization efficiency, with 95 % of the precursor splicing to produce circRNA. Optimization of spacer sequences flanking the circRNA payload eliminated interference from structured native exon sequences and enabled efficient circularization of RNAs up to 1,657 nt, including structured, repetitive, and naturally occurring sequences. Together, these results establish PCanPIE as a versatile and near-physiologically active addition to the group I intron PIE toolkit.

The eukaryotic ribosomal genes are multi-copy genes, transcribed from the rDNA, and approximately one third of them is actively transcribed in differentiated cells. A number of lncRNAs have been identified from the intergenic spacer between the rRNA genes, among those the spacer RNA and PAPAS that are involved silencing of rRNA gene copies by altering the chromatin configuration. Here, we have identified lncRNAs that are transcribed from the human rDNA loci and modulate the loci; IGS38 positively regulates rRNA gene transcription by associating to the 47S rRNA gene promoter and modulating the rRNA promoter accessibility while IGS32as associates with heterochromatin. IGS38 binds to the 47S gene promoter through the RNA pol I factors TAF1C and RRN3 as well as the Williams Syndrome Transcription Factor (WSTF), a component of the B-WICH chromatin remodelling complex. The increased accessibility of the promoter stabilises the architectural protein Upstream Binding Factor (UBF) at the rRNA promoter, thereby facilitating RNA pol I promoter escape. Furthermore, IGS38 knock down displays and increased dsRNA abundance in the cytoplasm with a weak induction of the dsRNA sensor OAS2, typically induced by interferon and viral dsRNA. Overall, the both IGS38 and IGS32as are chromatin associated lncRNAs involved in rDNA chromatin changes, and IGS38 is stimulating, together with WSTF, rRNA gene transcription in human cells. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=199 HEIGHT=200 SRC="FIGDIR/small/722362v1_ufig1.gif" ALT="Figure 1"> View larger version (29K): org.highwire.dtl.DTLVardef@14d4159org.highwire.dtl.DTLVardef@fd773forg.highwire.dtl.DTLVardef@a0030dorg.highwire.dtl.DTLVardef@1285301_HPS_FORMAT_FIGEXP M_FIG C_FIG IGS stabilises 47S rRNA transcription, disruption of IGS38 expression leads to the release of dsRNA in the cytoplasm and a weak immune activation of OAS2. Created by biorender (https://biorender.com/shortURL)

RNA-binding proteins (RBPs) play important roles in gene regulation. RNA editing-based approaches, such as TRIBE and STAMP, have gained wider use for identifying RNA targets of RBPs. These methods offer advantages over crosslinking-based approaches in terms of experimental simplicity and in vivo applicability. However, data analysis methods for these approaches remain underdeveloped, limiting sensitivity, and unbiased target prioritization. To address these limitations, we introduce SMARTIE (Systematic Machine-learning Approach for RBP Targets Identified by Editing), a machine-learning-based framework. SMARTIE robustly identifies and ranks RBP target RNAs from editing data by integrating statistical tests with replicate-aware and confidence-weighted features. Reanalysis of multiple published TRIBE datasets demonstrates the effectiveness of SMARTIE. It recovers targets of RBPs like Ataxin-2, TDP-43, Hrp48, Thor, GPATCH8, dFMRP and NonA. Notably, a model trained on TRIBE data generalizes to STAMP datasets, suggesting that SMARTIE learns universal signatures of editing-based RBP targeting there by enabling more accurate inference for RBP-RNA interactions. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=97 SRC="FIGDIR/small/726004v1_ufig1.gif" ALT="Figure 1"> View larger version (34K): org.highwire.dtl.DTLVardef@8b77e5org.highwire.dtl.DTLVardef@6c9416org.highwire.dtl.DTLVardef@6e33a5org.highwire.dtl.DTLVardef@100b7b5_HPS_FORMAT_FIGEXP M_FIG C_FIG

MicroRNAs are short noncoding RNAs that regulate gene expression and are commonly profiled by small RNA sequencing (miRNA-seq). Despite the widespread use of miRNA-seq, datasets are often analyzed with RNA-seq method such as DESeq2 or edgeR, which do not take into account the specific characteristics of miRNA-seq data. Here, we present a benchmark study of normalization and differential expression approaches using a realistic ground-truth dataset. By mixing mouse RNA of two organs, we generated expression trends while capturing biological and technical variability. Using monotonicity across the dataset and expected fold changes from the mixture design, we assessed normalization and differential expression methods. Normalization benchmarking showed that within-sample scaling, particularly Read Per Million (RPM), best preserved the expected monotonic trends, outperforming cross-sample methods such as TMM, rlog, and VST. These approaches sometimes recovered apparent monotonicity among abundant miRNAs, but inspection of individual profiles suggested likely over-correction. Regarding differential expression, edgeR consistently ranked among the best-performing methods across several metrics, including log2 fold-change estimation, with performance comparable to miRNA-seq-specific tools such as miRglmm and NBSR. DESeq2, edgeR-v4, and limma-based approaches tended to systematically underestimate log2 fold changes. Applying a common RPM-based normalization substantially improved the performance of cross-sample methods, highlighting the strong influence of normalization on differential expression analysis. Overall, our findings support within-sample scaling methods such as RPM for normalization, and edgeR, miRglmm, or NBSR for differential expression. The dataset has been made publicly available, providing a valuable resource for objective method comparison and future miRNA-seq software development.