Machine learning for RNA 2D structure prediction benchmarked on experimental data

Comparative analysis of RNA 3D structure prediction methods: towards enhanced modeling of RNA-ligand interactions

Accurate RNA structure models are crucial for designing small molecule ligands that modulate their functions. This study assesses six standalone RNA 3D structure prediction methods-DeepFoldRNA, RhoFold, BRiQ, FARFAR2, SimRNA and Vfold2, excluding web-based tools due to intellectual property concerns. We focus on reproducing the RNA structure existing in RNA-small molecule complexes, particularly on the ability to model ligand binding sites. Using a comprehensive set of RNA structures from the PDB, which includes diverse structural elements, we found that machine learning (ML)-based methods effectively predict global RNA folds but are less accurate with local interactions. Conversely, non-ML-based methods demonstrate higher precision in modeling intramolecular interactions, particularly with secondary structure restraints. Importantly, ligand-binding site accuracy can remain sufficiently high for practical use, even if the overall model quality is not optimal. With the recent release of AlphaFold 3, we included this advanced method in our tests. Benchmark subsets containing new structures, not used in the training of the tested ML methods, show that AlphaFold 3's performance was comparable to other ML-based methods, albeit with some challenges in accurately modeling ligand binding sites. This study underscores the importance of enhancing binding site prediction accuracy and the challenges in modeling RNA-ligand interactions accurately.

Ribonanza: deep learning of RNA structure through dual crowdsourcing

Prediction of RNA structure from sequence remains an unsolved problem, and progress has been slowed by a paucity of experimental data. Here, we present Ribonanza, a dataset of chemical mapping measurements on two million diverse RNA sequences collected through Eterna and other crowdsourced initiatives. Ribonanza measurements enabled solicitation, training, and prospective evaluation of diverse deep neural networks through a Kaggle challenge, followed by distillation into a single, self-contained model called RibonanzaNet. When fine tuned on auxiliary datasets, RibonanzaNet achieves state-of-the-art performance in modeling experimental sequence dropout, RNA hydrolytic degradation, and RNA secondary structure, with implications for modeling RNA tertiary structure.

State-of-the-RNArt: benchmarking current methods for RNA 3D structure prediction

RNAs are essential molecules involved in numerous biological functions. Understanding RNA functions requires the knowledge of their 3D structures. Computational methods have been developed for over two decades to predict the 3D conformations from RNA sequences. These computational methods have been widely used and are usually categorised as either ab initio or template-based. The performances remain to be improved. Recently, the rise of deep learning has changed the sight of novel approaches. Deep learning methods are promising, but their adaptation to RNA 3D structure prediction remains difficult. In this paper, we give a brief review of the ab initio, template-based and novel deep learning approaches. We highlight the different available tools and provide a benchmark on nine methods using the RNA-Puzzles dataset. We provide an online dashboard that shows the predictions made by benchmarked methods, freely available on the EvryRNA platform: https://evryrna.ibisc.univ-evry.fr/evryrna/state_of_the_rnart/.

sincFold: end-to-end learning of short- and long-range interactions in RNA secondary structure

MOTIVATION

Coding and noncoding RNA molecules participate in many important biological processes. Noncoding RNAs fold into well-defined secondary structures to exert their functions. However, the computational prediction of the secondary structure from a raw RNA sequence is a long-standing unsolved problem, which after decades of almost unchanged performance has now re-emerged due to deep learning. Traditional RNA secondary structure prediction algorithms have been mostly based on thermodynamic models and dynamic programming for free energy minimization. More recently deep learning methods have shown competitive performance compared with the classical ones, but there is still a wide margin for improvement.

RESULTS

In this work we present sincFold, an end-to-end deep learning approach, that predicts the nucleotides contact matrix using only the RNA sequence as input. The model is based on 1D and 2D residual neural networks that can learn short- and long-range interaction patterns. We show that structures can be accurately predicted with minimal physical assumptions. Extensive experiments were conducted on several benchmark datasets, considering sequence homology and cross-family validation. sincFold was compared with classical methods and recent deep learning models, showing that it can outperform the state-of-the-art methods.

Knotify_V2.0: Deciphering RNA Secondary Structures with H-Type Pseudoknots and Hairpin Loops

Accurately predicting the pairing order of bases in RNA molecules is essential for anticipating RNA secondary structures. Consequently, this task holds significant importance in unveiling previously unknown biological processes. The urgent need to comprehend RNA structures has been accentuated by the unprecedented impact of the widespread COVID-19 pandemic. This paper presents a framework, Knotify_V2.0, which makes use of syntactic pattern recognition techniques in order to predict RNA structures, with a specific emphasis on tackling the demanding task of predicting H-type pseudoknots that encompass bulges and hairpins. By leveraging the expressive capabilities of a Context-Free Grammar (CFG), the suggested framework integrates the inherent benefits of CFG and makes use of minimum free energy and maximum base pairing criteria. This integration enables the effective management of this inherently ambiguous task. The main contribution of Knotify_V2.0 compared to earlier versions lies in its capacity to identify additional motifs like bulges and hairpins within the internal loops of the pseudoknot. Notably, the proposed methodology, Knotify_V2.0, demonstrates superior accuracy in predicting core stems compared to state-of-the-art frameworks. Knotify_V2.0 exhibited exceptional performance by accurately identifying both core base pairing that form the ground truth pseudoknot in 70% of the examined sequences. Furthermore, Knotify_V2.0 narrowed the performance gap with Knotty, which had demonstrated better performance than Knotify and even surpassed it in Recall and F1-score metrics. Knotify_V2.0 achieved a higher count of true positives (tp) and a significantly lower count of false negatives (fn) compared to Knotify, highlighting improvements in Prediction and Recall metrics, respectively. Consequently, Knotify_V2.0 achieved a higher F1-score than any other platform. The source code and comprehensive implementation details of Knotify_V2.0 are publicly available on GitHub.

RNA3DB: A structurally-dissimilar dataset split for training and benchmarking deep learning models for RNA structure prediction

With advances in protein structure prediction thanks to deep learning models like AlphaFold, RNA structure prediction has recently received increased attention from deep learning researchers. RNAs introduce substantial challenges due to the sparser availability and lower structural diversity of the experimentally resolved RNA structures in comparison to protein structures. These challenges are often poorly addressed by the existing literature, many of which report inflated performance due to using training and testing sets with significant structural overlap. Further, the most recent Critical Assessment of Structure Prediction (CASP15) has shown that deep learning models for RNA structure are currently outperformed by traditional methods. In this paper we present RNA3DB, a dataset of structured RNAs, derived from the Protein Data Bank (PDB), that is designed for training and benchmarking deep learning models. The RNA3DB method arranges the RNA 3D chains into distinct groups (Components) that are non-redundant both with regard to sequence as well as structure, providing a robust way of dividing training, validation, and testing sets. Any split of these structurally-dissimilar Components are guaranteed to produce test and validations sets that are distinct by sequence and structure from those in the training set. We provide the RNA3DB dataset, a particular train/test split of the RNA3DB Components (in an approximate 70/30 ratio) that will be updated periodically. We also provide the RNA3DB methodology along with the source-code, with the goal of creating a reproducible and customizable tool for producing structurally-dissimilar dataset splits for structural RNAs.

RNA3DB: A structurally-dissimilar dataset split for training and benchmarking deep learning models for RNA structure prediction

With advances in protein structure prediction thanks to deep learning models like AlphaFold, RNA structure prediction has recently received increased attention from deep learning researchers. RNAs introduce substantial challenges due to the sparser availability and lower structural diversity of the experimentally resolved RNA structures in comparison to protein structures. These challenges are often poorly addressed by the existing literature, many of which report inflated performance due to using training and testing sets with significant structural overlap. Further, the most recent Critical Assessment of Structure Prediction (CASP15) has shown that deep learning models for RNA structure are currently outperformed by traditional methods. In this paper we present RNA3DB, a dataset of structured RNAs, derived from the Protein Data Bank (PDB), that is designed for training and benchmarking deep learning models. The RNA3DB method arranges the RNA 3D chains into distinct groups (Components) that are non-redundant both with regard to sequence as well as structure, providing a robust way of dividing training, validation, and testing sets. Any split of these structurally-dissimilar Components are guaranteed to produce test and validations sets that are distinct by sequence and structure from those in the training set. We provide the RNA3DB dataset, a particular train/test split of the RNA3DB Components (in an approximate 70/30 ratio) that will be updated periodically. We also provide the RNA3DB methodology along with the source-code, with the goal of creating a reproducible and customizable tool for producing structurally-dissimilar dataset splits for structural RNAs.

Effects of genetic variation on the structure of RNA and protein

Changes in the structure of RNA and protein, have an important impact on biological functions and are even important determinants of disease pathogenesis and treatment. Some genetic variations, including copy number variation, single nucleotide variation, and so on, can lead to changes in biological function and increased susceptibility to certain diseases by changing the structure of RNA or protein. With the development of structural biology and sequencing technology, a large amount of RNA and protein structure data and genetic variation data resources has emerged to be used to explain biological processes. Here, we reviewed the effects of genetic variation on the structure of RNAs and proteins, and investigated their impact on several diseases. An online resource (http://www.onethird-lab.com/gems/) to support convenient retrieval of common tools is also built. Finally, the challenges and future development of the effects of genetic variation on RNA and protein were discussed.

Concurrent prediction of RNA secondary structures with pseudoknots and local 3D motifs in an integer programming framework

MOTIVATION

The prediction of RNA structure canonical base pairs from a single sequence, especially pseudoknotted ones, remains challenging in a thermodynamic models that approximates the energy of the local 3D motifs joining canonical stems. It has become more and more apparent in recent years that the structural motifs in the loops, composed of noncanonical interactions, are essential for the final shape of the molecule enabling its multiple functions. Our capacity to predict accurate 3D structures is also limited when it comes to the organization of the large intricate network of interactions that form inside those loops.

RESULTS

We previously developed the integer programming framework RNA Motifs over Integer Programming (RNAMoIP) to reconcile RNA secondary structure and local 3D motif information available in databases. We further develop our model to now simultaneously predict the canonical base pairs (with pseudoknots) from base pair probability matrices with or without alignment. We benchmarked our new method over the all nonredundant RNAs below 150 nucleotides. We show that the joined prediction of canonical base pairs structure and local conserved motifs (i) improves the ratio of well-predicted interactions in the secondary structure, (ii) predicts well canonical and Wobble pairs at the location where motifs are inserted, (iii) is greatly improved with evolutionary information, and (iv) noncanonical motifs at kink-turn locations.

AVAILABILITY AND IMPLEMENTATION

The source code of the framework is available at https://gitlab.info.uqam.ca/cbe/RNAMoIP and an interactive web server at https://rnamoip.cbe.uqam.ca/.

A Hitchhiker's guide to RNA-RNA structure and interaction prediction tools

RNA biology has risen to prominence after a remarkable discovery of diverse functions of noncoding RNA (ncRNA). Most untranslated transcripts often exert their regulatory functions into RNA-RNA complexes via base pairing with complementary sequences in other RNAs. An interplay between RNAs is essential, as it possesses various functional roles in human cells, including genetic translation, RNA splicing, editing, ribosomal RNA maturation, RNA degradation and the regulation of metabolic pathways/riboswitches. Moreover, the pervasive transcription of the human genome allows for the discovery of novel genomic functions via RNA interactome investigation. The advancement of experimental procedures has resulted in an explosion of documented data, necessitating the development of efficient and precise computational tools and algorithms. This review provides an extensive update on RNA-RNA interaction (RRI) analysis via thermodynamic- and comparative-based RNA secondary structure prediction (RSP) and RNA-RNA interaction prediction (RIP) tools and their general functions. We also highlighted the current knowledge of RRIs and the limitations of RNA interactome mapping via experimental data. Then, the gap between RSP and RIP, the importance of RNA homologues, the relationship between pseudoknots, and RNA folding thermodynamics are discussed. It is hoped that these emerging prediction tools will deepen the understanding of RNA-associated interactions in human diseases and hasten treatment processes.

Using the structural diversity of RNA: protein interfaces to selectively target RNA with small molecules in cells: methods and perspectives

In recent years, RNA has gained traction both as a therapeutic molecule and as a therapeutic target in several human pathologies. In this review, we consider the approach of targeting RNA using small molecules for both research and therapeutic purposes. Given the primary challenge presented by the low structural diversity of RNA, we discuss the potential for targeting RNA: protein interactions to enhance the structural and sequence specificity of drug candidates. We review available tools and inherent challenges in this approach, ranging from adapted bioinformatics tools to in vitro and cellular high-throughput screening and functional analysis. We further consider two critical steps in targeting RNA/protein interactions: first, the integration of in silico and structural analyses to improve the efficacy of molecules by identifying scaffolds with high affinity, and second, increasing the likelihood of identifying on-target compounds in cells through a combination of high-throughput approaches and functional assays. We anticipate that the development of a new class of molecules targeting RNA: protein interactions to prevent physio-pathological mechanisms could significantly expand the arsenal of effective therapeutic compounds.

Integrating AlphaFold and deep learning for atomistic interpretation of cryo-EM maps

Interpretation of cryo-electron microscopy (cryo-EM) maps requires building and fitting 3D atomic models of biological molecules. AlphaFold-predicted models generate initial 3D coordinates; however, model inaccuracy and conformational heterogeneity often necessitate labor-intensive manual model building and fitting into cryo-EM maps. In this work, we designed a protein model-building workflow, which combines a deep-learning cryo-EM map feature enhancement tool, CryoFEM (Cryo-EM Feature Enhancement Model) and AlphaFold. A benchmark test using 36 cryo-EM maps shows that CryoFEM achieves state-of-the-art performance in optimizing the Fourier Shell Correlations between the maps and the ground truth models. Furthermore, in a subset of 17 datasets where the initial AlphaFold predictions are less accurate, the workflow significantly improves their model accuracy. Our work demonstrates that the integration of modern deep learning image enhancement and AlphaFold may lead to automated model building and fitting for the atomistic interpretation of cryo-EM maps.