Optimization of ribozyme activity using large-scale nucleic acid sequence data analysis by machine learning.

通过机器学习利用大规模核酸序列数据分析优化核酶活性。

• 批准号: 21J10391
• 负责人: ロッラッタナダムロン ラチャパン
• 金额: $ 0.96万
• 依托单位: Okinawa Institute of Science and Technology Graduate University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for JSPS Fellows
• 财政年份: 2021
• 资助国家: 日本
• 起止时间: 2021-04-28 至 2023-03-31
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-21J10391/
• 关键词: ribozymes; machine learning; deep learning; RNA engineering; Machine Learning; Fitness Landscape; Ribozyme; Evolution

The results from the directed evolution of ligase ribozymes augmented by supervised deep learning model have been published in a high impact peer reviewed journal and presented on an international conference. As a follow up work, I explored the use of unsupervised models to generate novel functional self-cleaving ribozymes. I trained three generative models on sequences of Twister self-cleaving ribozyme family. These computational works were conducted during my visiting position in Harvard Medical School. Experimental evaluation of these models showed promising preliminary results. Overall these works have shown that deep learning models can be used to design functional ribozymes sequences in both supervised and unsupervised manner.

由有监督的深度学习模型增强的连接酶核酶的定向进化的结果已经发表在一本高影响力的同行评议杂志上，并在一次国际会议上发表。作为后续工作，我探索了使用非监督模型来产生新的功能性自切割核酶。我针对Twister自裂解核酶家族的序列训练了三个繁殖模型。这些计算工作是我在哈佛医学院访问期间进行的。对这些模型的实验评估显示了有希望的初步结果。总体而言，这些工作表明，深度学习模型可以用来以监督和非监督的方式设计功能核酶序列。

项目成果

Experimental exploration of a ribozyme neutral network using evolutionary algorithm and deep learning.

• DOI: 10.1038/s41467-022-32538-z
• 发表时间: 2022-08-17
• 期刊: Nature communications
• 影响因子: 16.6