[论文解读] Leveraging Natural Language Processing to Unravel the Mystery of Life: A Review of NLP Approaches in Genomics, Transcriptomics, and Proteomics

Ella Rannon, David Burstein|ArXiv.org|Jun 2, 2025

Misinformation and Its Impacts被引用 3

一句话总结

本综述调查将NLP方法应用于生物序列，涵盖基因组、转录组和蛋白质组，从经典word2vec到transformer和基于hyena的模型，聚焦分词、架构及如结构预测和基因表达等应用。

ABSTRACT

Natural Language Processing (NLP) has transformed various fields beyond linguistics by applying techniques originally developed for human language to the analysis of biological sequences. This review explores the application of NLP methods to biological sequence data, focusing on genomics, transcriptomics, and proteomics. We examine how various NLP methods, from classic approaches like word2vec to advanced models employing transformers and hyena operators, are being adapted to analyze DNA, RNA, protein sequences, and entire genomes. The review also examines tokenization strategies and model architectures, evaluating their strengths, limitations, and suitability for different biological tasks. We further cover recent advances in NLP applications for biological data, such as structure prediction, gene expression, and evolutionary analysis, highlighting the potential of these methods for extracting meaningful insights from large-scale genomic data. As language models continue to advance, their integration into bioinformatics holds immense promise for advancing our understanding of biological processes in all domains of life.

研究动机与目标

• Motivate the use of NLP techniques for analysis of biological sequences across genomics, transcriptomics, and proteomics.
• Summarize how NLP methods are adapted to DNA, RNA, and protein data.
• Evaluate strengths and limitations of NLP approaches for different biological tasks.
• Discuss recent advances and applications in structure prediction, gene expression, and evolutionary analysis.
• Highlight future potential of integrating NLP with bioinformatics for large-scale genomics research.

提出的方法

• Survey of NLP methods adapted to biological sequences (DNA, RNA, proteins).
• Discussion of tokenization strategies and how they apply to biological data.
• Overview of model architectures, from classic word2vec to transformers and hyena operators.
• Evaluation of strengths, limitations, and suitability for various biological tasks.
• Synthesis of recent advances in structure prediction, gene expression, and evolutionary analysis.

实验结果

研究问题

• RQ1 How are NLP methods adapted to analyze DNA, RNA, and protein sequences across genomics, transcriptomics, and proteomics?
• RQ2 What tokenization strategies and model architectures are most effective for biological sequence data?
• RQ3 What are the strengths and limitations of NLP approaches for different biological tasks (e.g., structure prediction, gene expression, evolution)?
• RQ4 What are the recent advances enabling large-scale integration of NLP in bioinformatics?

主要发现

• NLP methods have evolved from classic word2vec to advanced transformer-based models and hyena operators for biological sequences.
• Tokenization strategies and model architectures critically affect performance on genomic tasks.
• Applications span structure prediction, gene expression analysis, and evolutionary studies.
• There is strong potential for NLP models to extract meaningful insights from large-scale genomic data.
• Integration of language-model approaches into bioinformatics is poised to advance understanding of biological processes across life.