RLALIGN: A Reinforcement Learning Approach for Multiple Sequence Alignment

Ramakrishnan, Ramchalam Kinattinkara; Singh, Jaspal; Blanchette, Mathieu

doi:10.1109/bibe.2018.00019

Cited by 7 publications

(8 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These steps are done by using a matrix that is a representation of each nucleotide and gap locus. In another paper [28], the A3C algorithm was employed to address the problem of speed in MSA. The scoring scheme used to optimize the SP score is the linear gap penalty approach, with a score of (-1, +1, -1) for (gap, match, mismatch).…”

Section: Applications From the Literaturementioning

confidence: 99%

See 1 more Smart Citation

Exploring Reinforcement Learning Methods for Multiple Sequence Alignment: A Brief Review

Gaad,

Chadi,

Sraitih

et al. 2023

BIO Web Conf.

View full text Add to dashboard Cite

Multiple sequence alignment (MSA) plays a vital role in uncovering similarities among biological sequences such as DNA, RNA, or proteins, providing valuable information about their structural, functional, and evolutionary relationships. However, MSA is a computationally challenging problem, with complexity growing exponentially as the number and length of sequences increase. Currently, standard MSA tools like ClustalW, T-Coffee, and MAFFT, which are based on heuristic algorithms, are widely used but still face many challenges due to the combinatorial explosion. Recent advancements in MSA algorithms have employed reinforcement learning (RL), particularly deep reinforcement learning (DRL), and demonstrated optimized execution time and accuracy with promising results. This is because deep reinforcement learning algorithms update their search policies using gradient descent, instead of exploring the entire solution space making it significantly faster and efficient. In this article, we provide an overview of the recent historical advancements in MSA algorithms, highlighting RL models used to tackle the MSA problem and main challenges and opportunities in this regard.

show abstract

Section: Applications From the Literaturementioning

confidence: 99%

“…Note: In Table 3 below, we included all papers presented previously that used RL for solving the MSA problem. However, for [28], where authors used the A3C algorithm, the metric results were not mentioned, therefore, we did not include it in Table 3.…”

Section: Benchmarkingmentioning

confidence: 99%

Exploring Reinforcement Learning Methods for Multiple Sequence Alignment: A Brief Review

Gaad,

Chadi,

Sraitih

et al. 2023

BIO Web Conf.

View full text Add to dashboard Cite

show abstract

“…Additionally, the choice of the aligners could be not totally exhaustive, due to practical reasons. Furthermore, we did not benchmark machine learning algorithms, such as DAVI [86], DeepFam [87], RLAlign [88] and DeepSF [89], namely neural networks and support vector machines, as well as the emerging developments in artificial intelligence, that represent the most emerging analytic field. Nevertheless, the major challenge is to deal with and interpret all the distinct output coming from each algorithm parameters.…”

Section: Limitationsmentioning

confidence: 99%

New evaluation methods of read mapping by 17 aligners on simulated and empirical NGS data: an updated comparison of DNA- and RNA-Seq data from Illumina and Ion Torrent technologies

Donato

Scimone

Rinaldi

et al. 2021

Neural Comput & Applic

View full text Add to dashboard Cite

During the last (15) years, improved omics sequencing technologies have expanded the scale and resolution of various biological applications, generating high-throughput datasets that require carefully chosen software tools to be processed. Therefore, following the sequencing development, bioinformatics researchers have been challenged to implement alignment algorithms for next-generation sequencing reads. However, nowadays selection of aligners based on genome characteristics is poorly studied, so our benchmarking study extended the “state of art” comparing 17 different aligners. The chosen tools were assessed on empirical human DNA- and RNA-Seq data, as well as on simulated datasets in human and mouse, evaluating a set of parameters previously not considered in such kind of benchmarks. As expected, we found that each tool was the best in specific conditions. For Ion Torrent single-end RNA-Seq samples, the most suitable aligners were CLC and BWA-MEM, which reached the best results in terms of efficiency, accuracy, duplication rate, saturation profile and running time. About Illumina paired-end osteomyelitis transcriptomics data, instead, the best performer algorithm, together with the already cited CLC, resulted Novoalign, which excelled in accuracy and saturation analyses. Segemehl and DNASTAR performed the best on both DNA-Seq data, with Segemehl particularly suitable for exome data. In conclusion, our study could guide users in the selection of a suitable aligner based on genome and transcriptome characteristics. However, several other aspects, emerged from our work, should be considered in the evolution of alignment research area, such as the involvement of artificial intelligence to support cloud computing and mapping to multiple genomes. Supplementary Information The online version contains supplementary material available at 10.1007/s00521-021-06188-z.

show abstract

“…Then Reza et al improved Mircea’s work and proposed an A3C (Asynchronous Advantage Actor Critic)-based method ( Jafari et al 2019 ). In addition, Ramakrishnan et al (2018) presented RLALIGN based on the A3C model in 2018. However, the above methods are only improved based on progressive alignment and do not use the characteristics of the sequence itself.…”

Section: Introductionmentioning

confidence: 99%

Multiple sequence alignment based on deep reinforcement learning with self-attention and positional encoding

Liu,

Yuan,

Zhang

et al. 2023

Bioinformatics

View full text Add to dashboard Cite

Motivation Multiple sequence alignments (MSA) is one of the hotspots of current research and is commonly used in sequence analysis scenarios. However, there is no lasting solution for MSA because it is an NP-complete (Nondeterministic Polynomially) problem, and the existing methods still have room to improve the accuracy. Result We propose DPAMSA (Deep reinforcement learning with Positional encoding and self-Attention for MSA), based on deep reinforcement learning (DRL), to enhance the accuracy of the alignment Specifically, inspired by the translation technique in natural language processing (NLP), we introduce self-attention and positional encoding to improve accuracy and reliability. Firstly, positional encoding encodes the position of the sequence to prevent the loss of nucleotide position information. Secondly, the self-attention model is used to extract the key features of the sequence. Then input the features into a multi-layer perceptron, which can calculate the insertion position of the gap according to the features. In addition, a novel reinforcement learning environment is designed to convert the classic progressive alignment into progressive column alignment, gradually generating each column’s sub-alignment. Finally, merge the sub-alignment into the complete alignment. Extensive experiments based on several datasets validate our method’s effectiveness for MSA, outperforming some state-of-the-art methods in terms of the Sum-of-pairs and Column scores. Availability and implementation The process is implemented in Python and available as open-source software from https://github.com/ZhangLab312/DPAMSA. Supplementary information Supplementary data are available at Bioinformatics online.

show abstract

RLALIGN: A Reinforcement Learning Approach for Multiple Sequence Alignment

Cited by 7 publications

References 21 publications

Exploring Reinforcement Learning Methods for Multiple Sequence Alignment: A Brief Review

Exploring Reinforcement Learning Methods for Multiple Sequence Alignment: A Brief Review

New evaluation methods of read mapping by 17 aligners on simulated and empirical NGS data: an updated comparison of DNA- and RNA-Seq data from Illumina and Ion Torrent technologies

Multiple sequence alignment based on deep reinforcement learning with self-attention and positional encoding

Contact Info

Product

Resources

About