蛋白质语言模型是偶然的 分类学家
Logan Hallee1,2, Tamar Peleg3, Nikolaos Rafailidis1
1Center for Bioinformatics and Computational Biology, University of Delaware.
bioRxiv : the preprint server for biology
|November 24, 2025
概括
蛋白质语言模型 (pLMs) 可以利用蛋白质-蛋白质相互作用 (PPI) 数据集中的物种差异,从而导致膨胀的性能. 仔细的数据策划对于可靠的计算生物学预测至关重要.
科学领域:
- 计算生物学 计算生物学
- 生物信息学是一种生物信息学.
- 机器学习在生物学中的应用
背景情况:
- 蛋白与蛋白相互作用 (PPI) 对生物功能至关重要,但在实验上很难确定.
- 计算方法,特别是蛋白质语言模型 (pLMs),为PPI预测提供可扩展的解决方案.
- 在多个物种数据集上的pLMs的高性能引发了关于底层预测机制的问题.
研究的目的:
- 为PPI预测引入和验证"偶然分类学家"假设.
- 调查pLM是否利用了基因信息而不是真正的相互作用特征.
- 提出减轻多个物种数据集中的这种混因素的策略.
主要方法:
- 用随机负样本分析蛋白质相互作用 (PPI) 数据集的分析.
- 使用蛋白质语言模型 (pLM) 嵌入来评估分类学相关性.
- 实施战略采样策略,用于同一物种内的负面实例.
- 在不同的数据采样策略中比较模型性能.
主要成果:
- 数据集中的真实PPI主要涉及来自同一物种的蛋白质,而负样本通常来自不同的物种.
- pLM嵌入可以准确地预测蛋白质对的分类起源,这表明它依赖于遗传学信号.
- 将负样本限制在同一物种对上显著降低了模型性能,证实了偶然的分类学效应.
- 策略性策划的多种模型优于单种模型,突出了管理良好的数据的潜力.
结论:
- "偶然的分类学家"是多种PPI预测中的一个重要的混者,模型利用物种标签而不是生物相互作用.
- 这种基因偏差并不仅限于PPI,也可能影响计算生物学中的其他监督学习任务.
- 仔细的数据策划和采样策略对于利用多种数据的好处来实现准确可靠的计算预测至关重要.
更多相关视频
03:14Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
1000
05:47Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems
Published on: June 13, 2025
1.3K
相关概念视频
Taxonomy
84.2K
Taxonomy is the science of defining and naming groups of biological organisms based on shared characteristics. It uses a hierarchy of increasingly inclusive categories with Latin names. The smallest units of taxonomy, species and genus, are used to assign a formal, taxonomic name to each species in a system. This classification system, referred to as binomial nomenclature, was formalized by Carolus Linnaeus in the 18th century.
Hierarchy of Taxonomy
The hierarchy that Carolus Linnaeus first...
Hierarchy of Taxonomy
The hierarchy that Carolus Linnaeus first...
84.2K
Genetic Lingo
113.6K
Overview
113.6K
The Tree of Life - Bacteria, Archaea, Eukaryotes
37.5K
The “tree of life” describes the evolution of life and the evolutionary relationships between organisms. The root of the tree is the common ancestor to all life on Earth. All other species radiate from this point, much like the branches of a tree. The numerous tips of these branches on the tree of life represent every living, or extant, species. Extinct species, which are species that no longer exist, can be found towards the center of the tree. Currently, these organisms, both...
37.5K
Phylogenetic Trees
49.1K
Phylogenetic trees come in many forms. It matters in which sequence the organisms are arranged from the bottom to the top of the tree, but the branches can rotate at their nodes without altering the information. The lines connecting individual nodes can be straight, angled, or even curved.
49.1K
Modern Molecular Taxonomy
554
Advancements in molecular biology have revolutionized the identification and characterization of bacteria, with multiple methods leveraging DNA sequencing for enhanced precision. As sequencing technologies improve and costs decline, these approaches are increasingly used in clinical, environmental, and evolutionary studies.Multilocus Sequence Typing (MLST) examines several housekeeping genes, essential chromosomal genes encoding cellular functions, to distinguish strains. Approximately...
554
Applications of Molecular Taxonomy
489
Molecular taxonomy has revolutionized the understanding and classification of bacteria, providing precise insights into their diversity, evolutionary relationships, and ecological roles. By utilizing molecular techniques such as DNA sequencing and fingerprinting, researchers have made significant strides in various fields related to bacterial studies.Resolving Taxonomic AmbiguitiesMolecular taxonomy has been instrumental in distinguishing closely related bacterial species initially thought to...
489
