⌘K
Change language Switch ThemeSign In
Narrow Mode
《自然通讯》重磅：分子之心 AI 技术解锁蛋白质设计新范式
量 量子位 @量子位的朋友们
One Sentence Summary
MoleculeMind, in collaboration with top-tier institutions, has published breakthrough results in Nature Communications. By leveraging the ComplexDDG algorithm, they doubled the insecticidal efficacy of scorpion toxin, ushering in a new 'mechanism analysis-AI design-experimental validation' paradigm for protein design.
Summary
This article reports on research published in the international journal Nature Communications by AI protein design leader MoleculeMind and its partners. By deeply integrating AI technology with molecular interaction mechanisms, the study successfully doubled the insecticidal efficacy of the scorpion toxin LqhαIT while maintaining low toxicity to mammals. The core breakthrough lies in the proprietary ComplexDDG algorithm, which can screen high-potential candidate molecules from a massive combinatorial space within hours, drastically shortening R&D cycles that previously took years. This study not only validates the potential of AI as an 'operating system for the bio-economy' in precision molecular engineering but also demonstrates its universal value in areas such as innovative drug development, industrial enzyme preparations, and eco-friendly materials. Currently, this technology has been integrated into their all-in-one platform, MoleculeOS.
Main Points
* 1. AI-driven protein design shifts from 'blind testing' to 'precision creation'.Traditional protein modification relies heavily on empirical trial-and-error with extremely low efficiency; AI algorithms can precisely design novel biomolecules with expected functions by analyzing molecular interaction mechanisms within vast combinatorial spaces. * 2. The ComplexDDG algorithm significantly enhances R&D efficiency and success rates.The algorithm designed 101 high-potential molecules from 20^66 combinations in just a few hours. Through multi-dimensional AI evaluation, 28 were selected for wet-lab experiments, ultimately achieving a doubling of insecticidal efficacy. * 3. Building a full-chain R&D closed loop of 'Mechanism-AI-Experiment'.This closed loop validates the universal value of AI protein design technology, which can be replicated in the design of complex biological macromolecules such as antibodies, vaccines, and novel enzymes, signaling a profound transformation in the bio-manufacturing industry.
Metadata
AI Score
88
Website qbitai.com
Published At Today
Length 2110 words (about 9 min)
Sign in to use highlight and note-taking features for a better reading experience. Sign in now
近日，AI蛋白质设计领域领军企业分子之心，携手天津大学、海南大学、南方科技大学、杜克大学等顶尖科研机构，在国际权威期刊《Nature Communications》上联合发表一项具有里程碑意义的研究成果。该研究深度融合AI蛋白质设计技术和分子作用机制解析，实现蝎毒素LqhαIT的杀虫效力翻倍。
更值得关注的是，这项研究构建起从机制发现到AI智能设计、再到实验验证的全链条研发闭环，强有力地验证了AI蛋白质设计技术作为生物经济的操作系统级技术，在革新生物大分子设计、加速新药研发等生物科技前沿创新中的巨大潜力与通用价值，标志着AI驱动的精准分子工程新范式的开启。
论文链接：https://www.nature.com/articles/s41467-026-70190-z AI蛋白设计：从“盲试”到“创造”的范式革新
自然界中，蛋白质作为承载生命功能的基石，是创新药物、诊断试剂、生物材料等各类生物功能分子的核心。
其中，海葵、蝎子等动物毒液中的多肽毒素，因其对昆虫靶标的高特异性和对哺乳动物的低毒性，不仅被视为新一代绿色生物农药的核心方向，也为新型多肽药物的开发提供了灵感。然而，蛋白质、多肽等生物分子的设计和应用长期受限于两大挑战：一是复杂生物分子的精细作用机制难以被彻底阐明，二是天然分子的功能改造与工程化优化高度依赖于传统的“经验试错、大海捞针”式筛选，研发投入巨大、周期漫长且成功率极低，成为制约药物研发、生物制造等领域创新的普遍瓶颈。
在本次研究中，联合团队完整揭示了绿海葵毒素Av3和蝎毒素LqhαIT这两种天然昆虫选择性毒素的作用机制，深入洞察了它们“精确识别与特异性作用”的核心分子密码，为后续该类分子的精准设计奠定了坚实基础，也为其他具有特定靶点选择性的生物分子设计提供了宝贵的通用方法论。
然而，即使有了机制蓝图，在海量组合空间中精确设计出具有预期功能的新型蛋白质或多肽依然是行业公认的巨大挑战。例如，由66个氨基酸组成的蝎毒素LqhαIT，其理论上的组合空间可达20的66次方种，若依靠传统实验筛选，其复杂程度堪比在浩瀚宇宙中寻找特定的一粒沙，成功率、效率极低。
分子之心自研ComplexDDG等AI蛋白质算法，成为突破这一行业普遍性瓶颈的关键钥匙。基于美洲大蠊钠通道NavPaS-蝎毒素LqhαIT复合物结构，该算法仅用数小时便设计出101个高潜力的候选分子。经AI算法的多维度智能评估，最终确定28个分子进入湿实验验证环节，并成功锁定一个杀虫效力翻倍，同时保留对哺乳动物低毒性的高效力候选分子。这一案例充分展现了AI蛋技术在复杂蛋白质精准设计上的强大能力。
ComplexDDG流程 AI智能引擎加速药物研发、生物制造万亿级产业革新
这套由AI深度驱动的研发体系，彻底颠覆了传统生物农药研发中高度依赖经验和漫无目的盲试的低效模式。它不仅大幅提升了蛋白质、多肽等生物大分子的研发效率与成功率，将原本动辄数年的研发周期缩短至数周甚至更短，更通过AI锁定真正具备高潜力的候选分子，极大提升研发成功率，为突破新药研发“双十定律”提供了颠覆性的解决方案。
除了在生物农药领域成功设计出杀虫效力翻倍的毒素，本次研究更为深远的意义在于，它成功跑通了“作用机制解析—AI精准设计—少量湿实验验证”的全链条创新闭环。这一实践不仅验证了AI蛋白质设计技术的强大应用价值，也为未来药物研发（如抗体、多肽、疫苗等）、新型酶制剂、生物材料等复杂生物大分子的设计提供了可复制的成功路径，预示创新药研发、生物制造乃至整个生物科技产业的深刻变革。
AI蛋白质设计技术，作为生物领域的下一代核心生产力工具，具备极强的通用价值和无限的创新潜力。它不仅能够从零开始设计自然界中不存在的新型生物分子，其应用前景更是极为广阔，将深远影响多个战略性产业：
在创新药物研发领域，分子之心正在借助AI设计出药效更优、靶向性更强、副作用更小的生物药，如长效双特异性抗体、减毒抗体、遮蔽肽、药物精准递送平台等，大幅加速新药研发进程，提升治疗效果，解决未满足的临床需求。
在工业酶制剂领域，分子之心AI已经设计出活性更高、在极端环境下（如高温、高压、强酸碱等）更稳定的新型酶，大幅提升生物催化效率，为精细化工、生物燃料等领域提供核心驱动力。
在生物材料与环境保护领域，分子之心正在借助AI设计可高效降解高分子材料、污染物的特异性酶，解锁更多环保解决方案，助力循环经济和可持续发展。
瞄准产业领域的真实需求，分子之心已构建起覆盖AI蛋白质预测、优化、从头设计全生命周期的完整技术体系。在复合物结构预测、蛋白质动态设计、蛋白质从头设计等底层核心技术上持续创新，保持全球领先水平。同时，面向产业关键场景，开发出生物药减毒设计、长效药物设计、耐极端环境蛋白设计等一系列高附加值专有技术平台。当前，分子之心已将这些底层算法和技术集成至自研的一站式AI蛋白质设计工业平台——“MoleculeOS”中，致力于用“AI按需设计”的前沿理念和方法，带动药物研发、新型酶制剂、绿色农业等领域创新提速，实现人类健康与可持续发展的未来。
转载来源：分子之心
本文为量子位获授权转载，观点仅为原作者所有。
量 量子位 @量子位的朋友们
One Sentence Summary
MoleculeMind, in collaboration with top-tier institutions, has published breakthrough results in Nature Communications. By leveraging the ComplexDDG algorithm, they doubled the insecticidal efficacy of scorpion toxin, ushering in a new 'mechanism analysis-AI design-experimental validation' paradigm for protein design.
Summary
This article reports on research published in the international journal Nature Communications by AI protein design leader MoleculeMind and its partners. By deeply integrating AI technology with molecular interaction mechanisms, the study successfully doubled the insecticidal efficacy of the scorpion toxin LqhαIT while maintaining low toxicity to mammals. The core breakthrough lies in the proprietary ComplexDDG algorithm, which can screen high-potential candidate molecules from a massive combinatorial space within hours, drastically shortening R&D cycles that previously took years. This study not only validates the potential of AI as an 'operating system for the bio-economy' in precision molecular engineering but also demonstrates its universal value in areas such as innovative drug development, industrial enzyme preparations, and eco-friendly materials. Currently, this technology has been integrated into their all-in-one platform, MoleculeOS.
Main Points
* 1. AI-driven protein design shifts from 'blind testing' to 'precision creation'.
Traditional protein modification relies heavily on empirical trial-and-error with extremely low efficiency; AI algorithms can precisely design novel biomolecules with expected functions by analyzing molecular interaction mechanisms within vast combinatorial spaces.
* 2. The ComplexDDG algorithm significantly enhances R&D efficiency and success rates.
The algorithm designed 101 high-potential molecules from 20^66 combinations in just a few hours. Through multi-dimensional AI evaluation, 28 were selected for wet-lab experiments, ultimately achieving a doubling of insecticidal efficacy.
* 3. Building a full-chain R&D closed loop of 'Mechanism-AI-Experiment'.
This closed loop validates the universal value of AI protein design technology, which can be replicated in the design of complex biological macromolecules such as antibodies, vaccines, and novel enzymes, signaling a profound transformation in the bio-manufacturing industry.
Key Quotes
* As an operating system-level technology for the bio-economy, AI protein design holds immense potential and universal value in revolutionizing the design of biological macromolecules and accelerating cutting-edge innovations in biotech, such as drug discovery. * The complexity is akin to finding a specific grain of sand in the vast universe, with extremely low success rates and efficiency. * This AI-driven R&D system completely overturns the inefficient model of traditional biopesticide development, which relies heavily on experience and aimless blind testing.
AI Score
88
Website qbitai.com
Published At Today
Length 2110 words (about 9 min)
Tags
AI Protein Design
MoleculeMind
ComplexDDG
Biotech
Nature Communications
Related Articles
* Yao Shunyu Lectures Face-to-Face with Tang Jie, Yang Zhilin, and Lin Junyang! Four Schema Heroes Debate Heroes at Zhongguancun * GPT-5.4 Released: OpenAI's First Unified Model, Truly Native * AI Starts to "Take Action", Alibaba's Qwen Leads the World * Google's New Paper Crashes Memory Stocks! 6x KV Cache Compression, Netizens: Silicon Valley Becomes Reality * Zhipu GLM-5 Technical Deep Dive: Full Compatibility with Domestic Chips Like Huawei Sparks Global Discussion and asynchronous reinforcement learning infrastructure, while completing full-stack adaptation for domestic chips.") HomeArticlesPodcastsVideosTweets