{"@id":"https://bioecon.ru/technology/deep-learning-biosynthetic-retrosynthesis/","@type":"Technology","cluster":"bioinformatics-omics","formats":{"html":"https://cn.bioecon.ru/technology/deep-learning-biosynthetic-retrosynthesis/","json":"https://cn.bioecon.ru/api/v1/technology/deep-learning-biosynthetic-retrosynthesis/index.json"},"id":"deep-learning-biosynthetic-retrosynthesis","industries":["bio-ai","research-tools"],"lang":"cn","links":{"industries":["https://bioecon.ru/industries/bio-ai/","https://bioecon.ru/industries/research-tools/"],"suppliers":["https://bioecon.ru/companies/enzymaster/","https://bioecon.ru/companies/schr-dinger/","https://bioecon.ru/companies/ginkgo-bioworks/","https://bioecon.ru/companies/insilico-medicine/","https://bioecon.ru/companies/mombak/","https://bioecon.ru/companies/previwo/","https://bioecon.ru/companies/protera/","https://bioecon.ru/companies/equinom/","https://bioecon.ru/companies/ardra-inc/","https://bioecon.ru/companies/atomwise/","https://bioecon.ru/companies/b-fab-gmbh/"]},"maturity":6,"name":"人工智能引导的罕见天然产物多步酶催化逆合成与途径生成技术","regions":["us","eu","cn"],"risk":"medium","section":"technology","sector":"synbio-enabling","sources":[{"date":"2026-11-15","entity":"pubs.acs.org","kind":"research","urls":["https://pubs.acs.org/doi/10.1021/acssynbio.9b00447"]},{"date":"2026-11-15","entity":"arxiv.org","kind":"web","urls":["https://arxiv.org/abs/2608.30702"]},{"date":"2026-11-15","entity":"ginkgo-bioworks","kind":"vendor","urls":[]},{"date":"2026-11-15","entity":"enzymaster","kind":"vendor","urls":[]},{"date":"2026-11-15","entity":"arzeda","kind":"vendor","urls":[]}],"stages":[{"n":1,"title":"目标分子化学拓扑逆向拆解"},{"n":2,"title":"全酶促生化反应空间知识检索"},{"n":3,"title":"吉布斯自由能热力学可行性打分"},{"n":4,"title":"自动化DNA元件装配与工程验证"}],"summary":"结合分子图神经网络与数百万已知生化反应知识图谱，针对复杂药用活性天然产物分子，全自动逆向推导最具热力学可行性与碳转化率的多酶催化级联路线，并精准推荐适配的工程酶序列与底物杂泛性改造位点。","url":"/technology/deep-learning-biosynthetic-retrosynthesis/","verified":"2026-11-15"}