通过参数高效微调,通过注入结构信息来增强蛋白质语言模型
Zixun Zhang1, Yuzhe Zhou1, Jiayou Zheng1
1FNii-Shenzhen, 2001 Longxiang Boulevard, Longgang District, Shenzhen, 518172, Guangdong, China; School of Science and Engineering, the Chinese University of Hong Kong (Shenzhen), 2001 Longxiang Boulevard, Longgang District, Shenzhen, 518172, Guangdong, China.
Computers in biology and medicine
|July 1, 2025
概括
结构信息注入调整 (SI-Tuning) 通过整合结构数据来增强蛋白质语言模型 (PLM). 这种方法可以提高下游任务的性能,例如有效地进行蛋白质分类和结合预测.
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.
- 结构生物学是结构生物学.
背景情况:
- 像ESM这样的大规模蛋白质语言模型 (PLM) 具有先进的蛋白质理解,但往往缺乏结构信息.
- 目前的PLM主要使用残留序列,限制它们捕获复杂蛋白质结构和功能的能力.
- 将结构数据集成到PLM中,为提高其在生物医学应用中的预测能力提供了重要机会.
研究的目的:
- 引入结构信息注入调整 (SI-Tuning),这是PLM的一个新型参数效率微调方法.
- 在不改变其核心参数的情况下,有效地将结构信息集成到现有的PLM中.
- 通过利用结构特征,提高PLM在各种下游任务的性能.
主要方法:
- SI-Tuning是一种参数高效的微调方法,可以保持预先训练的PLM参数被结.
- 针对特定任务的向量被优化为输入嵌入和注意力图.
- 结构特征,如二面角和距离图,用于导出这些向量来注入结构信息.
主要成果:
- 在650M ESM-2模型上,SI-Tuning在多个下游任务中显示出显著的性能改进.
- 在DeepLoc二进制分类上获得了93.95%的准确性,在金属离子结合上获得了76.05%的准确性.
- 超越了SaProt,一个具有集成结构建模的PLM,突出了SI-Tuning的有效性.
结论:
- SI-Tuning是一种有效的方法,可以通过将结构信息以参数有效的方式纳入PLM性能来提高.
- 该方法提供了相当大的计算效率,使其对广泛的生物医学应用具有价值.
- SI-Tuning代表了利用结构数据来改善蛋白质语言模型实用性的重大进步.
相关概念视频
02:07Improving 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
02:15Protein and Protein Structures
10.9K
10.9K
02:15Protein and Protein Structure
81.6K
Proteins are one of the most abundant organic molecules in living systems and have the most diverse range of functions of all macromolecules. Proteins may be structural, regulatory, contractile, or protective. They may serve in transport, storage, or membranes; or they may be toxins or enzymes. Their structures, like their functions, vary greatly. They are all, however, amino acid polymers arranged in a linear sequence.
A protein's shape is critical to its function. For example, an enzyme...
A protein's shape is critical to its function. For example, an enzyme...
81.6K
01:24Protein Organization
7.3K
Proteins are polymers of amino acid residues. They are versatile and responsible for different cellular functions, including DNA replication, molecular transport, catalysis, and structural support. Proteins have a hierarchical structure comprising at least three levels of organization: primary, secondary, and tertiary structure. Some large proteins have a quaternary structure where individual protein subunits are linked together.
The primary structure of a protein is its amino acid sequence....
The primary structure of a protein is its amino acid sequence....
7.3K
01:56Structural Protein Function
28.5K
Structural proteins are a category of proteins responsible for functions ranging from cell shape and movement to providing support to major structures such as bones, cartilage, hair, and muscles. This group includes proteins such as collagen, actin, myosin, and keratin.
Collagen, the most abundant protein in mammals, is found throughout the body. In connective tissue, such as skin, ligaments, and tendons, it provides tensile strength and elasticity. In bones and teeth, it mineralizes to...
Collagen, the most abundant protein in mammals, is found throughout the body. In connective tissue, such as skin, ligaments, and tendons, it provides tensile strength and elasticity. In bones and teeth, it mineralizes to...
28.5K
02:41Protein Complex Assembly
2.2K
2.2K