Tree Diet: Reducing the Treewidth to Unlock FPT Algorithms in RNA Bioinformatics

Marchand, Bertrand; Ponty, Yann; Bulteau, Laurent

doi:10.1101/2021.04.30.442158

Cited by 3 publications

(5 citation statements)

References 53 publications

(80 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can be seen that the treewidth rather correlates with the number of reticulation nodes than with network size (number of nodes and edges). Our study on 'real-world' phylogenetic networks suggests that treewidths are often low in practice; consequently Infrared can effectively compute network parsimony by solving the presented models for further algorithmic developments in bioinformatics [17,19,50]. Other previous work [3,20,46,48] could have directly profited from the Infrared framework.…”

Section: Discussionmentioning

confidence: 99%

“…In future work, we envision developing and making related methods accessible in a similar way. One exciting line would extend our work on tree decomposition-based automatic generation of dynamic programming schemes [19].…”

Section: Discussionmentioning

confidence: 99%

“…In parallel to the presented research, we used an early version of the system for original research in the area of negative RNA design (RNAPond, [17]). Other recent work has strong conceptual ties: Tree-Diet [18] (by using RNAPond and LicoRNA [3]) and AutoDP [19]). Moreover, as we show in this text, sequence and RNA sequencestructure alignment can be implemented following the models of LicoRNA [3] and [20]; both papers introduced closely related solving strategies for alignment.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Infrared: a declarative tree decomposition-powered framework for bioinformatics

Yao,

Marchand,

Berkemer

et al. 2024

Algorithms Mol Biol

View full text Add to dashboard Cite

Motivation Many bioinformatics problems can be approached as optimization or controlled sampling tasks, and solved exactly and efficiently using Dynamic Programming (DP). However, such exact methods are typically tailored towards specific settings, complex to develop, and hard to implement and adapt to problem variations. Methods We introduce the Infrared framework to overcome such hindrances for a large class of problems. Its underlying paradigm is tailored toward problems that can be declaratively formalized as sparse feature networks, a generalization of constraint networks. Classic Boolean constraints specify a search space, consisting of putative solutions whose evaluation is performed through a combination of features. Problems are then solved using generic cluster tree elimination algorithms over a tree decomposition of the feature network. Their overall complexities are linear on the number of variables, and only exponential in the treewidth of the feature network. For sparse feature networks, associated with low to moderate treewidths, these algorithms allow to find optimal solutions, or generate controlled samples, with practical empirical efficiency. Results Implementing these methods, the Infrared software allows Python programmers to rapidly develop exact optimization and sampling applications based on a tree decomposition-based efficient processing. Instead of directly coding specialized algorithms, problems are declaratively modeled as sets of variables over finite domains, whose dependencies are captured by constraints and functions. Such models are then automatically solved by generic DP algorithms. To illustrate the applicability of Infrared in bioinformatics and guide new users, we model and discuss variants of bioinformatics applications. We provide reimplementations and extensions of methods for RNA design, RNA sequence-structure alignment, parsimony-driven inference of ancestral traits in phylogenetic trees/networks, and design of coding sequences. Moreover, we demonstrate multidimensional Boltzmann sampling. These applications of the framework—together with our novel results—underline the practical relevance of Infrared. Remarkably, the achieved complexities are typically equivalent to the ones of specialized algorithms and implementations. Availability Infrared is available at https://amibio.gitlabpages.inria.fr/Infrared with extensive documentation, including various usage examples and API reference; it can be installed using Conda or from source.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Infrared: a declarative tree decomposition-powered framework for bioinformatics

Yao,

Marchand,

Berkemer

et al. 2024

Algorithms Mol Biol

View full text Add to dashboard Cite

show abstract

“…Pseudoknotted Structures Pseudoknots are an important type of base pairs that influence the function of an RNA (Staple & Butcher, 2005). Therefore, some approaches tried to design RNAs from pseudoknotted structures (Ren et al, 2005;Taneda, 2012;Kleinkauf et al, 2015;Merleau & Smerlak, 2022;Marchand et al, 2023). However, these algorithms work on a string notation in dot-bracket format (Hofacker et al, 1994), and thus, they cannot express base multiplets.…”

Section: Related Workmentioning

confidence: 99%

RNAinformer: Generative RNA Design With Tertiary Interactions

Patil,

Runge,

Franke

et al. 2024

Preprint

View full text Add to dashboard Cite

The design of RNAs that fulfill desired functions is one of the major challenges in computational biology. The function of an RNA molecule depends on its structure and a strong structure-to-function relationship is already achieved on the secondary structure level of RNA. Therefore, computational RNA design is often interpreted as the inversion of a folding algorithm: Given a target secondary structure, find an RNA sequence that folds into the desired structure. However, existing RNA design approaches cannot invert state-of-the-art folding algorithms because they can only predict a limited set of base interactions. In this work, we propose RNAinformer, a novel generative transformer based approach to the inverse RNA folding problem. Leveraging axial attention, we are able to process secondary structures represented as adjacency matrices, which allows us to invert state-of-the-art folding algorithms. Consequently, RNAinformer is the first model capable of designing RNAs from secondary structures without base pair restrictions. We demonstrate RNAinformer's strong performance across different RNA design benchmarks and showcase its novelty by inverting a state-of-the-art deep learning based secondary structure prediction algorithm.

show abstract

“…The Zuker-style dynamic programming approach to fold an RNA sequence of length L without pseudoknots requires time complexity in O(L 3 ) [1] [2]. Algorithms that take into account pseudoknots are even more complex and have been reported to require significantly more computational power ranging from O(L 4 ) to O(L 6 ) [3] [4], or higher [5].…”

Section: Introductionmentioning

confidence: 99%

Automatic recognition of complementary strands: Lessons regarding machine learning abilities in RNA folding

Chasles

Major

2023

Preprint

View full text Add to dashboard Cite

Prediction of RNA secondary structure from single sequences still needs substantial improvements. The application of machine learning (ML) to this problem has become increasingly popular. However, ML algorithms are prone to overfitting, limiting the ability to learn more about the inherent mechanisms governing RNA folding. It is natural to use high-capacity models when solving such a difficult task, but poor generalization is expected when too few examples are available. Here, we report the relation between capacity and performance on a simpler related problem: determining whether two sequences are fully complementary. Our analysis focused on the impact of model architecture and capacity as well as dataset size and nature on classification accuracy. We observed that low-capacity models are better suited for learning with mislabelled training examples, while large capacities improve the ability to generalize to structurally dissimilar data. Given a more complex task like RNA folding, it comes as no surprise that the scarcity of usable examples hurdles the applicability of machine learning techniques to this field.

show abstract

Tree Diet: Reducing the Treewidth to Unlock FPT Algorithms in RNA Bioinformatics

Cited by 3 publications

References 53 publications

Infrared: a declarative tree decomposition-powered framework for bioinformatics

Infrared: a declarative tree decomposition-powered framework for bioinformatics

RNAinformer: Generative RNA Design With Tertiary Interactions

Automatic recognition of complementary strands: Lessons regarding machine learning abilities in RNA folding

Contact Info

Product

Resources

About