Recent Deep Learning Methodology Development for RNA–RNA Interaction Prediction

Fang, Yi; Pan, Xiaoyong; Shen, Hong‐Bin

doi:10.3390/sym14071302

Cited by 7 publications

(7 citation statements)

References 97 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…miRNAs play crucial roles in RNA silencing and post-transcriptional regulation of gene expression by specifically targeting certain mRNAs for degradation and translational repression(https://en.wikipedia.org/wiki/Mi-croRNA). 29 miRNA genes are typically transcribed by RNA polymerase II (Pol II), which binds to a promoter located near the DNA sequence, encoding what will become the hairpin loop of a precursor (pre)-miRNA. Pre-miRNAs are approximately 70 nucleotides long and fold into imperfect stem-loop structures.…”

Section: Algebraic Geometry Of Mirnasmentioning

confidence: 99%

Group Theory of Messenger RNA Metabolism and Disease

Planat,

Amaral,

Chester

et al. 2024

Gene Expr

View full text Add to dashboard Cite

Our recent work has focused on the application of infinite group theory and related algebraic geometric tools in the context of transcription factors and microRNAs. We were able to differentiate between "healthy" nucleotide sequences and disrupted sequences that may be associated with various diseases. In this paper, we extend our efforts to the study of messenger RNA (mRNA) metabolism, showcasing the power of our approach. We investigate (1) mRNA translation in prokaryotes and eukaryotes, (2) polyadenylation in eukaryotes, which is crucial for nuclear export, translation, stability, and splicing of mRNA, (3) microRNAs involved in RNA silencing and post-transcriptional regulation of gene expression, and (4) identification of disrupted sequences that could lead to potential illnesses. To achieve this, we used: (a) infinite (finitely generated) groups f p , with generators representing the r + 1 distinct nucleotides and a relation between them [e.g., the consensus sequence in the mRNA translation (i), the poly (A) tail in item (ii), and the microRNA seed in item (iii)]; (b) aperiodicity theory, which connects healthy groups f p to free groups F r of rank r and their profinite completion ˆr F , and (c) the representation theory of groups f p over the space-time-spin group SL 2 (C), highlighting the role of surfaces with isolated singularities in the character variety. Our approach could potentially contribute to the understanding of the molecular mechanisms underlying various diseases and help develop new diagnostic or therapeutic strategies.

show abstract

Section: Algebraic Geometry Of Mirnasmentioning

confidence: 99%

Group Theory of Messenger RNA Metabolism and Disease

Planat,

Amaral,

Chester

et al. 2024

Gene Expr

View full text Add to dashboard Cite

show abstract

“…This task maps two RNA sequences, x a of length L 1 and x b of length L 2 , to binary labels 0/1, where 0 indicates no interaction between the two RNA sequences and 1 indicates interaction. Accurate prediction of RNA-RNA interactions can provide valuable insights into RNA-mediated regulatory mechanisms and enhance our understanding of biological processes, including gene expression, splicing and translation 45 .…”

Section: Articlementioning

confidence: 99%

Multi-purpose RNA language modelling with motif-aware pretraining and type-guided fine-tuning

Wang,

Bian,

et al. 2024

Nat Mach Intell

View full text Add to dashboard Cite

Pretrained language models have shown promise in analysing nucleotide sequences, yet a versatile model excelling across diverse tasks with a single pretrained weight set remains elusive. Here we introduce RNAErnie, an RNA-focused pretrained model built upon the transformer architecture, employing two simple yet effective strategies. First, RNAErnie enhances pretraining by incorporating RNA motifs as biological priors and introducing motif-level random masking in addition to masked language modelling at base/subsequence levels. It also tokenizes RNA types (for example, miRNA, lnRNA) as stop words, appending them to sequences during pretraining. Second, subject to out-of-distribution tasks with RNA sequences not seen during the pretraining phase, RNAErnie proposes a type-guided fine-tuning strategy that first predicts possible RNA types using an RNA sequence and then appends the predicted type to the tail of sequence to refine feature embedding in a post hoc way. Our extensive evaluation across seven datasets and five tasks demonstrates the superiority of RNAErnie in both supervised and unsupervised learning. It surpasses baselines with up to 1.8% higher accuracy in classification, 2.2% greater accuracy in interaction prediction and 3.3% improved F1 score in structure prediction, showcasing its robustness and adaptability with a unified pretrained foundation.

show abstract

“…MicroRNAs (miRNAs) are important players of the expression and regulation of genes by targeting specific messenger RNAs (mRNAs) for degradation or translational repression [38,39]. The miRNAs are approximately 22 nt (where nt is for nucleotide) long single-stranded RNA molecules.…”

Section: Algebraic Morphology Of Micrornasmentioning

confidence: 99%

Algebraic Morphology of DNA–RNA Transcription and Regulation

2023

View full text Add to dashboard Cite

Transcription factors (TFs) and microRNAs (miRNAs) are co-actors in genome-scale decoding and regulatory networks, often targeting common genes. To discover the symmetries and invariants of the transcription and regulation at the scale of the genome, in this paper, we introduce tools of infinite group theory and of algebraic geometry to describe both TFs and miRNAs. In TFs, the generator of the group is a DNA-binding domain while, in miRNAs, the generator is the seed of the sequence. For such a generated (infinite) group π, we compute the SL(2,C) character variety, where SL(2,C) is simultaneously a ‘space-time’ (a Lorentz group) and a ‘quantum’ (a spin) group. A noteworthy result of our approach is to recognize that optimal regulation occurs when π looks similar to a free group Fr (r=1 to 3) in the cardinality sequence of its subgroups, a result obtained in our previous papers. A non-free group structure features a potential disease. A second noteworthy result is about the structure of the Groebner basis G of the variety. A surface with simple singularities (such as the well known Cayley cubic) within G is a signature of a potential disease even when π looks similar to a free group Fr in its structure of subgroups. Our methods apply to groups with a generating sequence made of two to four distinct DNA/RNA bases in {A,T/U,G,C}. We produce a few tables of human TFs and miRNAs showing that a disease may occur when either π is away from a free group or G contains surfaces with isolated singularities.

show abstract

Recent Deep Learning Methodology Development for RNA–RNA Interaction Prediction

Cited by 7 publications

References 97 publications

Group Theory of Messenger RNA Metabolism and Disease

Group Theory of Messenger RNA Metabolism and Disease

Multi-purpose RNA language modelling with motif-aware pretraining and type-guided fine-tuning

Algebraic Morphology of DNA–RNA Transcription and Regulation

Contact Info

Product

Resources

About