深度学习模型用于基于无偏序的PPI预测高原,准确度为0.65
Timo Reim1,2, Anne Hartebrodt2, David B Blumenthal2
1Data Science in Systems Biology, TUM School of Life Sciences, Technical University of Munich, Freising, 85354, Germany.
Bioinformatics (Oxford, England)
|July 15, 2025
概括
使用基于序列的方法预测蛋白质与蛋白质相互作用是具有挑战性的,因为数据泄露. 虽然ESM-2嵌入可以提高性能,但可靠的预测可能需要结构数据.
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.
- 蛋白质相互作用分析分析
背景情况:
- 蛋白与蛋白相互作用 (PPI) 对细胞功能至关重要.
- 准确的PPI计算预测仍然是一个重大挑战.
- 之前的评估方案和数据泄露问题已经掩盖了基于序列的PPI预测的进展.
研究的目的:
- 研究蛋白质嵌入对基于序列的PPI预测的影响.
- 评估不同模型架构和嵌入策略的性能.
- 为了确定基于序列的模型是否可以隐式学习PPI预测的联系地图.
主要方法:
- 利用ESM-2蛋白质嵌入用于基于序列的PPI预测.
- 与不同复杂度的模型进行比较,每种蛋白质和每种令牌嵌入.
- 评估了自我注意力和交叉注意力机制的影响.
- 分析了模型学习接触地图作为中间表示的能力.
主要成果:
- 无论模型架构如何,ESM-2嵌入在PPI预测中显著解释了性能增长.
- 所有测试的基于序列的模型均以0.65.5的精度稳定.
- 基于序列的模型不能隐含地学习联系人地图.
- 性能增长归因于嵌入,而不是固有的基于序列的学习能力.
结论:
- 像ESM-2这样的蛋白质嵌入是基于序列的PPI预测最近性能改进的关键驱动因素.
- 目前基于序列的模型有局限性,不能隐式学习联系地图.
- 结构信息可能对于实现可靠和准确的PPI预测至关重要.
相关概念视频
Protein-protein Interfaces
13.4K
Many proteins form complexes to carry out their functions, making protein-protein interactions (PPIs) essential for an organism's survival. Most PPIs are stabilized by numerous weak noncovalent chemical forces. The physical shape of the interfaces determines the way two proteins interact. Many globular proteins have closely-matching shapes on their surfaces, which form a large number of weak bonds. Additionally, many PPIs occur between two helices or between a surface cleft and a...
13.4K
Protein Networks
4.1K
An organism can have thousands of different proteins, and these proteins must cooperate to ensure the health of an organism. Proteins bind to other proteins and form complexes to carry out their functions. Many proteins interact with multiple other proteins creating a complex network of protein interactions.
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
These interactions can be represented through maps depicting protein-protein interaction networks, represented as nodes and edges. Nodes are circles that are representative of a protein,...
4.1K
Conserved Binding Sites
4.4K
Many proteins’ biological role depends on their interactions with their ligands, small molecules that bind to specific locations on the protein known as ligand-binding sites. Ligand-binding sites are often conserved among homologous proteins as these sites are critical for protein function.
Binding sites are often located in large pockets, and if their location on a protein’s surface is unknown, it can be predicted using various approaches. The energetic method computationally...
Binding sites are often located in large pockets, and if their location on a protein’s surface is unknown, it can be predicted using various approaches. The energetic method computationally...
4.4K
Protein-Protein Interfaces
3.8K
3.8K
Conservation of Protein Domains Over Different Proteins
11.4K
Protein domains are small structurally independent units that are part of a single amino acid chain. Although these domains are often structurally independent, they may rely on synergistic effects to perform their functions as part of a larger protein. Protein domains may be conserved within the same organism, as well as across different organisms.
A limited set of protein domains often duplicate and recombine during evolution. These domains can be organized in different combinations to...
A limited set of protein domains often duplicate and recombine during evolution. These domains can be organized in different combinations to...
11.4K
Protein Folding Quality Check in the RER
3.8K
ER is the primary site for the maturation and folding of soluble and transmembrane secretory proteins. The calnexin cycle is a specific chaperone system that folds and assesses the confirmation of N-glycosylated proteins before they can exit the ER lumen. The primary players of this quality check pipeline are the lectins, ER-resident chaperones, and a glucosyl transferase enzyme. In case the calnexin system in the lumen fails to salvage a misfolded protein, it is transported to the cytoplasm...
3.8K


