Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

Microbial dysbiosis and inferred functional profiling reveals the potential role of <i>Methylobacterium</i> in prostate cancer.

Frontiers in cellular and infection microbiology·2026
Same author

Depletion of Neurocan in the Prefrontal Cortex Impairs Temporal Order Recognition, Cognitive Flexibility and Perisomatic GABAergic Innervation.

Cellular and molecular neurobiology·2026
Same author

Artificial intelligence directed computational protein design: lessons from COVID-19 for pandemic-ready vaccines and antibody therapeutics.

Journal of pharmacy & pharmaceutical sciences : a publication of the Canadian Society for Pharmaceutical Sciences, Societe canadienne des sciences pharmaceutiques·2026
Same author

Ensemble threshold Boolean modeling reveals robust attractors and regulatory drivers in pediatric leukemia.

Computers in biology and medicine·2026
Same author

Epidemiology and Clinical Profile of Brachial Plexus Injuries in Semi-Urban Punjab: Insights From a Retrospective Study at a Tertiary Care Center.

Cureus·2026
Same author

Pharmacological targeting of the NLRP3 LRR domain with isothiazolinones overcomes CRID3-resistant inflammation.

EMBO molecular medicine·2026

相关实验视频

Updated: Jun 1, 2025

Protein Crystallization for X-ray Crystallography
09:27

Protein Crystallization for X-ray Crystallography

Published on: January 16, 2011

63.4K

对蛋白质结晶的蛋白质语言模型进行基准测试.

Raghvendra Mall1, Rahul Kaushik2, Zachary A Martinez3

  • 1Biotechnology Research Center, Technology Innovation Institute, P.O. Box 9639, Abu Dhabi, United Arab Emirates. raghvendra.mall@tii.ae.

Scientific reports
|January 18, 2025
PubMed
概括

我们对开放的蛋白质语言模型 (PLMs) 进行了基准测试,以预测蛋白质结晶. 嵌入ESM2模型显著提高了预测准确性,优于现有方法,并允许设计新的可结晶蛋白质.

关键词:
基准测试 (benchmarking) 是一种比较的方法.开放的蛋白质语言模型 (PLM)蛋白质结晶的过程中的蛋白质结晶.蛋白质的产生是蛋白质的产生.

更多相关视频

Author Spotlight: High-Throughput Screening to Obtain Crystal Hits for Protein Crystallography
06:19

Author Spotlight: High-Throughput Screening to Obtain Crystal Hits for Protein Crystallography

Published on: March 10, 2023

4.2K
Automated Protocols for Macromolecular Crystallization at the MRC Laboratory of Molecular Biology
11:20

Automated Protocols for Macromolecular Crystallization at the MRC Laboratory of Molecular Biology

Published on: January 24, 2018

16.3K

相关实验视频

Last Updated: Jun 1, 2025

Protein Crystallization for X-ray Crystallography
09:27

Protein Crystallization for X-ray Crystallography

Published on: January 16, 2011

63.4K
Author Spotlight: High-Throughput Screening to Obtain Crystal Hits for Protein Crystallography
06:19

Author Spotlight: High-Throughput Screening to Obtain Crystal Hits for Protein Crystallography

Published on: March 10, 2023

4.2K
Automated Protocols for Macromolecular Crystallization at the MRC Laboratory of Molecular Biology
11:20

Automated Protocols for Macromolecular Crystallization at the MRC Laboratory of Molecular Biology

Published on: January 24, 2018

16.3K

科学领域:

  • 结构生物学 结构生物学
  • 计算生物学 计算生物学
  • 机器学习 机器学习

背景情况:

  • 蛋白质结构的确定对于理解功能至关重要,通常依赖于X射线晶体学.
  • 目前正在开发in silico方法,特别是深度学习,以从序列中预测蛋白质结晶倾向,解决实验限制.
  • 蛋白语言模型 (PLM) 为学习基于序列的表示提供了一个强大的方法.

研究的目的:

  • 为了对开放蛋白语言模型 (PLMs) 的性能进行基准测试,以预测蛋白质结晶倾向.
  • 为此任务确定最有效的PLM和机器学习分类器.
  • 探索PLM在产生新型,可能结晶的蛋白质中的实用性.

主要方法:

  • 在平均蛋白质嵌入上使用LightGBM/XGBoost分类器对各种开放的PLM (ESM2,Ankh,ProtT5-XL,ProstT5,xTrimoPGLM,SaProt) 进行基准测试.
  • 将基于PLM的方法与最新的基于序列的预测器 (DeepCrystal,ATTCrys,CLPred) 进行比较.
  • 微调 ProtGPT2 模型以进行 de novo 蛋白质生成,然后进行多阶段过.

主要成果:

  • 使用ESM2嵌入式的LightGBM分类器 (30/36层,150M/3B参数) 与所有其他测试模型相比,在多个指标 (AUPR,AUC,F1) 上显示出显著的性能增长 (高达3倍).
  • 该研究确定了最有效的PLM来预测蛋白质结晶的结果.
  • 通过精心调整的ProtGPT2模型和严格的过,成功设计和识别了五种新型,潜在的结晶蛋白.

结论:

  • 开放的蛋白质语言模型,特别是ESM2,在预测蛋白质结晶倾向方面表现出卓越的性能.
  • TRILL平台有效地使PLM用于这一关键任务的使用实现了民主化.
  • 基于PLM的生成方法有望加速发现具有理想性质的新型蛋白质,例如结晶性.