Search

Messenger RNA-based medicines hold immense potential, as evidenced by their rapid deployment as COVID-19 vaccines. However, worldwide distribution of mRNA molecules has been limited by their thermostability, which is fundamentally limited by the intrinsic instability of RNA molecules to a chemical degradation reaction called in-line hydrolysis. Predicting the degradation of an RNA molecule is a key task in designing more stable RNA-based therapeutics. Here, we describe a crowdsourced machine learning competition ("Stanford OpenVaccine") on Kaggle, involving single-nucleotide resolution measurements on 6043 102-130-nucleotide diverse RNA constructs that were themselves solicited through crowdsourcing on the RNA design platform Eterna. The entire experiment was completed in less than 6 months, and 41% of nucleotide-level predictions from the winning model were within experimental error of the ground truth measurement. Furthermore, these models generalized to blindly predicting orthogonal degradation data on much longer mRNA molecules (504-1588 nucleotides) with improved accuracy compared to previously published models. Top teams integrated natural language processing architectures and data augmentation techniques with predictions from previous dynamic programming models for RNA secondary structure. These results indicate that such models are capable of representing in-line hydrolysis with excellent accuracy, supporting their use for designing stabilized messenger RNAs. The integration of two crowdsourcing platforms, one for data set creation and another for machine learning, may be fruitful for other urgent problems that demand scientific discovery on rapid timescales.

Solving the RNA inverse folding problem is a critical prerequisite to RNA design, an emerging field in bioengineering with a broad range of applications from reaction catalysis to cancer therapy. Although significant progress has been made in developing machine-based inverse RNA folding algorithms, current approaches still have difficulty designing sequences for large or complex targets. On the other hand, human players of the online RNA design game EteRNA have consistently shown superior performance in this regard, being able to readily design sequences for targets that are challenging for machine algorithms. Here we present a novel approach to the RNA design problem, SentRNA, a design agent consisting of a fully-connected neural network trained end-to-end using human-designed RNA sequences. We show that through this approach, SentRNA can solve complex targets previously unsolvable by any machine-based approach and achieve state-of-the-art performance on two separate challenging test sets. Our results demonstrate that incorporating human design strategies into a design algorithm can significantly boost machine performance and suggests a new paradigm for machine-based RNA design., Comment: 27 pages (not including Supplementary Information), 9 figures, 7 tables

The three-dimensional conformations of non-coding RNAs underpin their biochemical functions but have largely eluded experimental characterization. Here, we report that integrating a classic mutation/rescue strategy with high-throughput chemical mapping enables rapid RNA structure inference with unusually strong validation. We revisit a paradigmatic 16S rRNA domain for which SHAPE (selective 2`-hydroxyl acylation with primer extension) suggested a conformational change between apo- and holo-ribosome conformations. Computational support estimates, data from alternative chemical probes, and mutate-and-map (M2) experiments expose limitations of prior methodology and instead give a near-crystallographic secondary structure. Systematic interrogation of single base pairs via a high-throughput mutation/rescue approach then permits incisive validation and refinement of the M2-based secondary structure and further uncovers the functional conformation as an excited state (25+/-5% population) accessible via a single-nucleotide register shift. These results correct an erroneous SHAPE inference of a ribosomal conformational change and suggest a general mutate-map-rescue approach for dissecting RNA dynamic structure landscapes.

To facilitate the analysis of large-scale high-throughput capillary electrophoresis data, we previously proposed a suite of efficient analysis software named HiTRACE (High Throughput Robust Analysis of Capillary Electrophoresis). HiTRACE has been used extensively for quantitating data from RNA and DNA structure mapping experiments, including mutate-and-map contact inference, chromatin footprinting, the EteRNA RNA design project and other high-throughput applications. However, HiTRACE is based on a suite of command-line MATLAB scripts that requires nontrivial efforts to learn, use, and extend. Here we present HiTRACE-Web, an online version of HiTRACE that includes standard features previously available in the command-line version as well as additional features such as automated band annotation and flexible adjustment of annotations, all via a user-friendly environment. By making use of parallelization, the on-line workflow is also faster than software implementations available to most users on their local computers. Free access: http://hitrace.org

Chemical mapping methods probe RNA structure by revealing and leveraging correlations of a nucleotide's structural accessibility or flexibility with its reactivity to various chemical probes. Pioneering work by Lucks and colleagues has expanded this method to probe hundreds of molecules at once on an Illumina sequencing platform, obviating the use of slab gels or capillary electrophoresis on one molecule at a time. Here, we describe optimizations to this method from our lab, resulting in the MAP-seq protocol (Multiplexed Accessibility Probing read out through sequencing), version 1.0. The protocol permits the quantitative probing of thousands of RNAs at once, by several chemical modification reagents, on the time scale of a day using a table-top Illumina machine. This method and a software package MAPseeker (http://simtk.org/home/map_seeker) address several potential sources of bias, by eliminating PCR steps, improving ligation efficiencies of ssDNA adapters, and avoiding problematic heuristics in prior algorithms. We hope that the step-by-step description of MAP-seq 1.0 will help other RNA mapping laboratories to transition from electrophoretic to next-generation sequencing methods and to further reduce the turnaround time and any remaining biases of the protocol., Comment: 22 pages, 5 figures

The Rosetta molecular modeling software package provides experimentally tested and rapidly evolving tools for the 3D structure prediction and high-resolution design of proteins, nucleic acids, and a growing number of non-natural polymers. Despite its free availability to academic users and improving documentation, use of Rosetta has largely remained confined to developers and their immediate collaborators due to the code's difficulty of use, the requirement for large computational resources, and the unavailability of servers for most of the Rosetta applications. Here, we present a unified web framework for Rosetta applications called ROSIE (Rosetta Online Server that Includes Everyone). ROSIE provides (a) a common user interface for Rosetta protocols, (b) a stable application programming interface for developers to add additional protocols, (c) a flexible back-end to allow leveraging of computer cluster resources shared by RosettaCommons member institutions, and (d) centralized administration by the RosettaCommons to ensure continuous maintenance. This paper describes the ROSIE server infrastructure, a step-by-step 'serverification' protocol for use by Rosetta developers, and the deployment of the first nine ROSIE applications by six separate developer teams: Docking, RNA de novo, ERRASER, Antibody, Sequence Tolerance, Supercharge, Beta peptide design, NCBB design, and VIP redesign. As illustrated by the number and diversity of these applications, ROSIE offers a general and speedy paradigm for serverification of Rosetta applications that incurs negligible cost to developers and lowers barriers to Rosetta use for the broader biological community. ROSIE is available at http://rosie.rosettacommons.org.

Chemical mapping is a widespread technique for structural analysis of nucleic acids in which a molecule's reactivity to different probes is quantified at single-nucleotide resolution and used to constrain structural modeling. This experimental framework has been extensively revisited in the past decade with new strategies for high-throughput read-outs, chemical modification, and rapid data analysis. Recently, we have coupled the technique to high-throughput mutagenesis. Point mutations of a base-paired nucleotide can lead to exposure of not only that nucleotide but also its interaction partner. Carrying out the mutation and mapping for the entire system gives an experimental approximation of the molecules contact map. Here, we give our in-house protocol for this mutate-and-map strategy, based on 96-well capillary electrophoresis, and we provide practical tips on interpreting the data to infer nucleic acid structure., Comment: 22 pages, 5 figures

Consistently predicting biopolymer structure at atomic resolution from sequence alone remains a difficult problem, even for small sub-segments of large proteins. Such loop prediction challenges, which arise frequently in comparative modeling and protein design, can become intractable as loop lengths exceed 10 residues and if surrounding side-chain conformations are erased. This article introduces a modeling strategy based on a 'stepwise ansatz', recently developed for RNA modeling, which posits that any realistic all-atom molecular conformation can be built up by residue-by-residue stepwise enumeration. When harnessed to a dynamic-programming-like recursion in the Rosetta framework, the resulting stepwise assembly (SWA) protocol enables enumerative sampling of a 12 residue loop at a significant but achievable cost of thousands of CPU-hours. In a previously established benchmark, SWA recovers crystallographic conformations with sub-Angstrom accuracy for 19 of 20 loops, compared to 14 of 20 by KIC modeling with a comparable expenditure of computational power. Furthermore, SWA gives high accuracy results on an additional set of 15 loops highlighted in the biological literature for their irregularity or unusual length. Successes include cis-Pro touch turns, loops that pass through tunnels of other side-chains, and loops of lengths up to 24 residues. Remaining problem cases are traced to inaccuracies in the Rosetta all-atom energy function. In five additional blind tests, SWA achieves sub-Angstrom accuracy models, including the first such success in a protein/RNA binding interface, the YbxF/kink-turn interaction in the fourth RNA-puzzle competition. These results establish all-atom enumeration as a systematic approach to protein structure that can leverage high performance computing and physically realistic energy functions to more consistently achieve atomic resolution., Comment: Identity of four-loop blind test protein and parts of figures 5 have been omitted in this preprint to ensure confidentiality of the protein structure prior to its public release

For decades, dimethyl sulfate (DMS) mapping has informed manual modeling of RNA structure in vitro and in vivo. Here, we incorporate DMS data into automated secondary structure inference using a pseudo-energy framework developed for 2'-OH acylation (SHAPE) mapping. On six non-coding RNAs with crystallographic models, DMS- guided modeling achieves overall false negative and false discovery rates of 9.5% and 11.6%, comparable or better than SHAPE-guided modeling; and non-parametric bootstrapping provides straightforward confidence estimates. Integrating DMS/SHAPE data and including CMCT reactivities give small additional improvements. These results establish DMS mapping - an already routine technique - as a quantitative tool for unbiased RNA structure modeling.

Chemical purity of RNA samples is critical for high-precision studies of RNA folding and catalytic behavior, but such purity may be compromised by photodamage accrued during ultraviolet (UV) visualization of gel-purified samples. Here, we quantitatively assess the breadth and extent of such damage by using reverse transcription followed by single-nucleotide-resolution capillary electrophoresis. We detected UV-induced lesions across a dozen natural and artificial RNAs including riboswitch domains, other non-coding RNAs, and artificial sequences; across multiple sequence contexts, dominantly at but not limited to pyrimidine doublets; and from multiple lamps that are recommended for UV shadowing in the literature. Most strikingly, irradiation time-courses reveal detectable damage within a few seconds of exposure, and these data can be quantitatively fit to a 'skin effect' model that accounts for the increased exposure of molecules near the top of irradiated gel slices. The results indicate that 200-nucleotide RNAs subjected to 20 seconds or less of UV shadowing can incur damage to 20% of molecules, and the molecule-by-molecule distribution of these lesions is more heterogeneous than a Poisson distribution. Photodamage from UV shadowing is thus likely a widespread but unappreciated cause of artifactual heterogeneity in quantitative and single-molecule-resolution RNA biophysical measurements.

The tertiary structures of functional RNA molecules remain difficult to decipher. A new generation of automated RNA structure prediction methods may help address these challenges but have not yet been experimentally validated. Here we apply four prediction tools to a remarkable class of double glycine riboswitches that exhibit ligand-binding cooperativity. A novel method (BPPalign), RMdetect, JAR3D, and Rosetta 3D modeling give consistent predictions for a new stem P0 and kink-turn motif. These elements structure the linker between the RNAs' double aptamers. Chemical mapping on the F. nucleatum riboswitch with SHAPE, DMS, and CMCT probing, mutate-and-map studies, and mutation/rescue experiments all provide strong evidence for the structured linker. Under solution conditions that separate two glycine binding transitions, disrupting this helix-junction-helix structure gives 120-fold and 6- to 30-fold poorer association constants for the two transitions, corresponding to an overall energetic impact of 4.3 \pm 0.5 kcal/mol. Prior biochemical and crystallography studies from several labs did not include this critical element due to over-truncation of the RNA. We argue that several further undiscovered elements are likely to exist in the flanking regions of this and other RNA switches, and automated prediction tools can now play a powerful role in their detection and dissection.

Three-dimensional RNA models fitted into crystallographic density maps exhibit pervasive conformational ambiguities, geometric errors and steric clashes. To address these problems, we present enumerative real-space refinement assisted by electron density under Rosetta (ERRASER), coupled to Python-based hierarchical environment for integrated 'xtallography' (PHENIX) diffraction-based refinement. On 24 data sets, ERRASER automatically corrects the majority of MolProbity-assessed errors, improves the average Rfree factor, resolves functionally important discrepancies in noncanonical structure and refines low-resolution models to better match higher-resolution models.

We have established an RNA Mapping Database (RMDB) to enable a new generation of structural, thermodynamic, and kinetic studies from quantitative single-nucleotide-resolution RNA structure mapping (freely available at http://rmdb.stanford.edu). Chemical and enzymatic mapping is a rapid, robust, and widespread approach to RNA characterization. Since its recent coupling with high-throughput sequencing techniques, accelerated software pipelines, and large-scale mutagenesis, the volume of mapping data has greatly increased, and there is a critical need for a database to enable sharing, visualization, and meta-analyses of these data. Through its on-line front-end, the RMDB allows users to explore single-nucleotide-resolution chemical accessibility data in heat-map, bar-graph, and colored secondary structure graphics; to leverage these data to generate secondary structure hypotheses; and to download the data in standardized and computer-friendly files, including the RDAT and community-consensus SNRNASM formats. At the time of writing, the database houses 38 entries, describing 2659 RNA sequences and comprising 355,084 data points, and is growing rapidly., Comment: 20 pages, 2 figures

Atomic-accuracy structure prediction of macromolecules is a long-sought goal of computational biophysics. Accurate modeling should be achievable by optimizing a physically realistic energy function but is presently precluded by incomplete sampling of a biopolymer's many degrees of freedom. We present herein a working hypothesis, called the "stepwise ansatz", for recursively constructing well-packed atomic-detail models in small steps, enumerating several million conformations for each monomer and covering all build-up paths. By implementing the strategy in Rosetta and making use of high-performance computing, we provide first tests of this hypothesis on a benchmark of fifteen RNA loop modeling problems drawn from riboswitches, ribozymes, and the ribosome, including ten cases that were not solvable by prior knowledge based modeling approaches. For each loop problem, this deterministic stepwise assembly (SWA) method either reaches atomic accuracy or exposes flaws in Rosetta's all-atom energy function, indicating the resolution of the conformational sampling bottleneck. To our knowledge, SWA is the first enumerative, ab initio build-up method to systematically outperform existing Monte Carlo and knowledge-based methods for 3D structure prediction. As a rigorous experimental test, we have applied SWA to a small RNA motif of previously unknown structure, the C7.2 tetraloop/tetraloop-receptor, and stringently tested this blind prediction with nucleotide-resolution structure mapping data.

Motivation: Capillary electrophoresis (CE) of nucleic acids is a workhorse technology underlying high-throughput genome analysis and large-scale chemical mapping for nucleic acid structural inference. Despite the wide availability of CE-based instruments, there remain challenges in leveraging their full power for quantitative analysis of RNA and DNA structure, thermodynamics, and kinetics. In particular, the slow rate and poor automation of available analysis tools have bottlenecked a new generation of studies involving hundreds of CE profiles per experiment. Results: We propose a computational method called high-throughput robust analysis for capillary electrophoresis (HiTRACE) to automate the key tasks in large-scale nucleic acid CE analysis, including the profile alignment that has heretofore been a rate-limiting step in the highest throughput experiments. We illustrate the application of HiTRACE on thirteen data sets representing 4 different RNAs, three chemical modification strategies, and up to 480 single mutant variants; the largest data sets each include 87,360 bands. By applying a series of robust dynamic programming algorithms, HiTRACE outperforms prior tools in terms of alignment and fitting quality, as assessed by measures including the correlation between quantified band intensities between replicate data sets. Furthermore, while the smallest of these data sets required 7 to 10 hours of manual intervention using prior approaches, HiTRACE quantitation of even the largest data sets herein was achieved in 3 to 12 minutes. The HiTRACE method therefore resolves a critical barrier to the efficient and accurate analysis of nucleic acid structure in experiments involving tens of thousands of electrophoretic bands., Comment: Revised to include Supplement. Availability: HiTRACE is freely available for download at http://hitrace.stanford.edu

Non-coding RNA molecules fold into precise base pairing patterns to carry out critical roles in genetic regulation and protein synthesis. We show here that coupling systematic mutagenesis with high-throughput SHAPE chemical mapping enables accurate base pair inference of domains from ribosomal RNA, ribozymes, and riboswitches. For a six-RNA benchmark that challenged prior chemical/computational methods, this mutate-and-map strategy gives secondary structures in agreement with crystallographic data (2 % error rates), including a blind test on a double-glycine riboswitch. Through modeling of partially ordered RNA states, the method enables the first test of an 'interdomain helix-swap' hypothesis for ligand-binding cooperativity in a glycine riboswitch. Finally, the mutate-and-map data report on tertiary contacts within non-coding RNAs; coupled with the Rosetta/FARFAR algorithm, these data give nucleotide-resolution three-dimensional models (5.7 {\AA} helix RMSD) of an adenine riboswitch. These results highlight the promise of a two-dimensional chemical strategy for inferring the secondary and tertiary structures that underlie non-coding RNA behavior.

Single-nucleotide-resolution chemical mapping for structured RNA is being rapidly advanced by new chemistries, faster readouts, and coupling to computational algorithms. Recent tests have shown that selective 2'-hydroxyl acylation by primer extension (SHAPE) can give near-zero error rates (0-2%) in modeling the helices of RNA secondary structure. Here, we benchmark the method using six molecules for which crystallographic data are available: tRNA(phe) and 5S rRNA from Escherichia coli, the P4-P6 domain of the Tetrahymena group I ribozyme, and ligand-bound domains from riboswitches for adenine, cyclic di-GMP, and glycine. SHAPE-directed modeling of these highly structured RNAs gave an overall false negative rate (FNR) of 17% and a false discovery rate (FDR) of 21%, with at least one helix prediction error in five of the six cases. Extensive variations of data processing, normalization, and modeling parameters did not significantly mitigate modeling errors. Only one varation, filtering out data collected with deoxyinosine triphosphate during primer extension, gave a modest improvement (FNR = 12%, and FDR = 14%). The residual structure modeling errors are explained by the insufficient information content of these RNAs' SHAPE data, as evaluated by a nonparametric bootstrapping analysis. Beyond these benchmark cases, bootstrapping suggests a low level of confidence (<50%) in the majority of helices in a previously proposed SHAPE-directed model for the HIV-1 RNA genome. Thus, SHAPE-directed RNA modeling is not always unambiguous, and helix-by-helix confidence estimates, as described herein, may be critical for interpreting results from this powerful methodology., Comment: Biochemistry, Article ASAP (Aug. 15, 2011)

A complete macromolecule modeling package must be able to solve the simplest structure prediction problems. Despite recent successes in high resolution structure modeling and design, the Rosetta software suite fares poorly on deceptively small protein and RNA puzzles, some as small as four residues. To illustrate these problems, this manuscript presents extensive Rosetta results for four well-defined test cases: the 20-residue mini-protein Trp cage, an even smaller disulfide-stabilized conotoxin, the reactive loop of a serine protease inhibitor, and a UUCG RNA tetraloop. In contrast to previous Rosetta studies, several lines of evidence indicate that conformational sampling is not the major bottleneck in modeling these small systems. Instead, approximations and omissions in the Rosetta all-atom energy function currently preclude discriminating experimentally observed conformations from de novo models at atomic resolution. These molecular "puzzles" should serve as useful model systems for developers wishing to make foundational improvements to this powerful modeling suite., Comment: Published in PLoS One as a manuscript for the RosettaCon 2010 Special Collection

No existing algorithm can start with arbitrary RNA sequences and return the precise, three-dimensional structures that ensures their biological function. This chapter outlines current algorithms for automated RNA structure prediction (including our own FARNA-FARFAR), highlights their successes, and dissects their limitations, using a tetraloop and the sarcin/ricin motif as examples. The barriers to future advances are considered in light of three particular challenges: improving computational sampling, reducing reliance on experimentally solved structures, and avoiding coarse-grained representations of atomic-level interactions. To help meet these challenges and better understand the current state of the field, we propose an ongoing community-wide CASP-style experiment for evaluating the performance of current structure prediction algorithms., Comment: K. Beauchamp & P. Sripakdeevong are equally contributing authors. Submission for book: RNA 3D Structure Analysis and Prediction, editors: N. Leontis & E. Westhof

We present the results of radio, optical and near-infrared observations of the field of TOC J0233.3+3021, a cluster of milliJansky radiosources from the TexOx Cluster survey. In an observation of this field with the Ryle Telescope (RT) at 15 GHz, we measure a decrement in the cosmic microwave background (CMB) of $-675 \pm 95 \mu$Jy on the RT's $\approx$ 0.65 k$\lambda$ baseline. Using optical and infrared imaging with the McDonald 2.7-m Smith Reflector, Calar Alto 3.5-m telescope and UKIRT, we identify the host galaxies of five of the radiosources and measure magnitudes of $R \approx 24$, $J \approx 20$, $K \approx 18$. The CMB decrement is consistent with the Sunyaev-Zel'dovich (SZ) effect of a massive cluster of galaxies, which if modelled as a spherical King profile of core radius $\theta_C = 20^{\prime\prime}$ has a central temperature decrement of $900 \mu$K. The magnitudes and colours of the galaxies are consistent with those of old ellipticals at $z \sim 1$. We therefore conclude that TOC J0233.3+3021 is a massive, high redshift cluster. These observations add to the growing evidence for a significant population of massive clusters at high redshift, and demonstrate the effectiveness of combining searches for AGN `signposts' to clusters with the redshift-independence of the SZ effect., Comment: Six pages; accepted for publication in MNRAS. Version with full-resolution UV plot available from http://www.mrao.cam.ac.uk/~garret/MB185.ps

We describe a maximum-likelihood technique for the removal of contaminating radio sources from interferometric observations of the Sunyaev-Zel'dovich (SZ) effect. This technique, based on a simultaneous fit for the radio sources and extended SZ emission, is also compared to techniques previously applied to Ryle Telescope observations and is found to be robust. The technique is then applied to new observations of the cluster Abell 611, and a decrement of -540 +/- 125 microJy/beam is found. This is combined with a ROSAT HRI image and a published ASCA temperature to give an Hubble constant estimate of 52+24-16 km/s/Mpc., Comment: 8 pages, 9 figures. Submitted to MNRAS

We present new Ryle Telescope (RT) observations of the Sunyaev Zel'dovich (SZ) decrement from the cluster Abell 773. The field contains a number of faint radio sources that required careful subtraction. We use ASCA observations to measure the gas temperature and a ROSAT HRI image to model the gas distribution. Normalising the gas distribution to fit the RT visibilities returns a value of H_0 of 77 (+19,-15) km/s/Mpc (1-sigma errors) for an Einstein-de-Sitter universe, or 85 (+20,-17) km/s/Mpc for a flat model with Omega_Lambda = 0.7. The errors quoted include estimates of the effects of the principal errors: noise in the SZ measurement, gas temperature uncertainty, and line-of sight depth uncertainty., Comment: 5 pages, 6 figures, MNRAS in press (minor revisions)