预测RNA结构利用预训语言模型的注意力
Ioannis Papazoglou1,2, Alexios Chatzigoulas1, George Tsekenis1
1Biomedical Research Foundation, Academy of Athens, Athens 11527, Greece.
Journal of chemical information and modeling
|July 2, 2025
概括
目前的人工智能语言模型由于架构限制而难以准确预测RNA二级和三级结构,尽管它们具有诊断和治疗的潜力.
科学领域:
- 计算生物学是一种计算生物学.
- 生物信息学是一种生物信息学.
- 分子生物学分子生物学
背景情况:
- 非编码RNA的功能角色与其二级和三级结构密切相关.
- 准确的RNA结构预测对于开发新型诊断和治疗方法至关重要.
- 预测三维 (3D) RNA结构仍然是一个重要的计算挑战.
研究的目的:
- 评估预训练核酸语言模型在预测RNA二级和三级结构方面的有效性.
- 评估人工智能 (AI) 和大型语言模型 (LLM) 在RNA结构预测中的潜力.
- 确定目前用于RNA结构分析的AI方法的局限性.
主要方法:
- 预训练模型的评估:RNABERT,ERNIE-RNA,RNA基础模型 (RNA-FM),Ribo核酸语言模型 (RiNALMo) 和DNABERT.
- 对二级和三级RNA结构预测任务的模型性能评估.
- 分析影响模型准确性的架构约束.
主要成果:
- 目前的核酸语言模型表明,在RNA中捕获基本结构信息的能力有限.
- 现有模型中的架构约束阻碍了有效的RNA二级和三级结构预测.
- 虽然LLMs的应用很有希望,但需要进一步开发以克服当前的局限性.
结论:
- 现有的预训练的核酸语言模型还不适合准确的RNA结构预测.
- 对新型人工智能架构的进一步研究是必要的,以提高RNA结构预测能力.
- 解决架构限制是释放人工智能在基于RNA的诊断和治疗方面的潜力的关键.
相关概念视频
Nucleic Acid Structure
7.2K
The pentose sugar in DNA is deoxyribose, while in RNA the pentose sugar is ribose. The difference between the sugars is the presence of the hydroxyl group on the ribose's second carbon and a hydrogen on the deoxyribose's second carbon. The phosphate residue attaches to the hydroxyl group of the 5′ carbon of one sugar and the hydroxyl group of the 3′ carbon of the sugar of the next nucleotide, which forms a 5′ to 3′ phosphodiester linkage.
DNA Structure
DNA...
DNA Structure
DNA...
7.2K
RNA Stability
33.9K
Intact DNA strands can be found in fossils, while scientists sometimes struggle to keep RNA intact under laboratory conditions. The structural variations between RNA and DNA underlie the differences in their stability and longevity. Because DNA is double-stranded, it is inherently more stable. The single-stranded structure of RNA is less stable but also more flexible and can form weak internal bonds. Additionally, most RNAs in the cell are relatively short, while DNA can be up to 250 million...
33.9K
Types of RNA
6.5K
Three main types of RNA are involved in protein synthesis: messenger RNA (mRNA), transfer RNA (tRNA), and ribosomal RNA (rRNA). These RNAs perform diverse functions and can be broadly classified as protein-coding or non-coding RNA. Non-coding RNAs play important roles in regulating gene expression in response to developmental and environmental changes. Non-coding RNAs in prokaryotes can be manipulated to develop more effective antibacterial drugs for human or animal use.
RNA Performs Diverse...
RNA Performs Diverse...
6.5K
pre-mRNA Processing
53.6K
In eukaryotic cells, transcripts made by RNA polymerase are modified and processed before exiting the nucleus. Unprocessed RNA is called precursor mRNA or pre-mRNA to distinguish it from mature mRNA.
Once about 20-40 ribonucleotides have been joined together by RNA polymerase, a group of enzymes adds a “cap” to the 5’ end of the growing transcript. In this process, a 5’ phosphate is replaced by modified guanosine that has a methyl group attached to it (7-Methyl...
Once about 20-40 ribonucleotides have been joined together by RNA polymerase, a group of enzymes adds a “cap” to the 5’ end of the growing transcript. In this process, a 5’ phosphate is replaced by modified guanosine that has a methyl group attached to it (7-Methyl...
53.6K
Chromatin Structure Regulates pre-mRNA Processing
7.2K
In eukaryotic cells, nascent mRNA transcripts need to undergo many post-transcriptional modifications to reach the cell cytoplasm and translate into functional proteins. For a long time, transcription and pre-mRNA processing were considered two independent events that occur sequentially in the cell. However, it has now been well established that transcription and pre-mRNA processing are two simultaneous processes that are precisely regulated inside the cell.
The chromatin structure, especially...
The chromatin structure, especially...
7.2K
Improving Translational Accuracy
11.9K
Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...
11.9K


