超算上百亿年找不到的答案，百度这个算法十分钟内找到了最优解

{"type":"doc","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"blockquote","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":10}}],"text":"作者 | 百度研究院"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":10}}],"text":"编辑 | 刘燕"}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"疫情期间，mRNA疫苗以更高的安全性和有效性，更快的研发和生产速度被寄予厚望，成为疫情防控的重要武器，也是各国生命科学领域的关键竞争点。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"虽然目前mRNA技术已被成功商业化，但全球的科研团队仍面临一个统一难题：mRNA疫苗普遍稳定性低，致使它非常“脆弱”，很容易因为降解而失效，这成为mRNA疫苗研发过程中的巨大阻碍。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"mRNA的这种不稳定性给mRNA疫苗的存储、运输和免疫原性都带来了挑战，这些问题的解决与mRNA疫苗序列的设计息息相关。目前业内普遍认可的破解之道，就是设计出具有稳定结构且蛋白质翻译效率高的mRNA序列。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"虽然找到了解决路径，但是要想破局并不容易。序列设计所需要的庞大的计算量，成为研发过程中最严峻的“拦路虎”。与新冠病毒刺突蛋白对应的mRNA，若采用最传统的遍历方法去寻找一条稳定的序列，需要查看10^632次方个mRNA序列。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"打个比方，假设一台超级计算机每秒计算一个序列，在宇宙诞生到现在的138亿年时间里，可能连潜在序列亿万分之一都无法搜索完成。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"面对如此庞大的搜索空间，必须要有全新的算法，才能在有限的时间内寻找到那条最稳定有效的疫苗序列。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"早在2018年，百度研究院就开展了 RNA 二级结构领域的研究。得益于这几年在生物计算领域的研究积累，在疫情之初，百度研究院做到了快速响应，并在2020年4月正式对外发布了LinearDesign mRNA序列优化算法。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"与依靠穷举法搜索稳定mRNA序列的思路不同，LinearDesign通过动态规划算法，将序列稳定性和蛋白质翻译效率指标进行联合优化，可在10分钟内找到比天然序列更加稳定且蛋白质表达水平更高的新冠mRNA疫苗序列，实现了在最短时间内用最高的效率得出最优化的方案。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"为了进一步验证LinearDesign算法的实际有效性，百度联合mRNA药物公司斯微生物进行了一系列实验研究。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"本次实验研究基于LinearDesign算法设计的七条编码新冠病毒刺突蛋白的mRNA疫苗序列，对这些mRNA分子的稳定性、编码蛋白表达水平以及动物免疫原性等方面开展了深入研究，并与用传统方法设计的基准mRNA序列进行了头对头的对比。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"近日，百度和斯微生物联合公布了mRNA疫苗序列设计算法LinearDesign的新冠病毒疫苗生物实验结果：\t在稳定性、蛋白质表达水平以及免疫原性等多个衡量疫苗的重要指标上，LinearDesign设计的新冠疫苗序列均大幅优于传统方法设计的基准序列，尤其在抗原中和抗体滴度这一衡量疫苗有效性的关键指标上，LinearDesign设计的序列最高超过基准序列达到20倍，表示LinearDesign设计的疫苗能够有效产生免疫保护；其实际有效性和应用于生物制药领域的价值被充分验证。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/4f\/ab\/4ff4a73dfc46abb3ab6a0bb9f349e8ab.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":"center","origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":10}}],"text":"百度LinearDesign算法设计的七条疫苗序列（A-G）以及基准序列（H）等相关信息"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"百度LinearDesign算法从理论层面和生物学实验层面得到有效性验证，为将AI应用于生命科学探索出一条实际可行的道路。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"LinearDesign算法也进一步验证了AI+制药的价值。这项技术还具有广泛的适用性，除了应用于新冠病毒mRNA疫苗的研发，还能包括传染病疫苗、肿瘤疫苗、单抗等各种疫苗和药物的研发，也进一步验证了人工智能、生物计算技术在生命科学领域的实际应用价值。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"以LinearDesign算法为代表的生物计算技术，将大大缩短新药研发的周期、降低研发成本，并能提高医疗诊断的准确性和效率。"}]}]}