Computational studies of in vitro selection of RNAs

描述(申请人提供)：项目摘要/摘要。通过体外选择实验发现的合成RNA在生物医学科学和生物技术中有广泛的应用，包括抑制蛋白质功能的治疗性适配子和控制基因表达的核酶。然而，体外选择RNA的过程缺乏系统的计算分析，可能会增加发现复杂合成RNA的可能性。为了填补这一重大缺口，我们开发了用于设计结构化RNA池和优化RNA功能的计算方法，以提高体外选择和定向进化实验的生产率。体外选择实验和我们的计算分析表明，设计的具有不同结构基序的RNA池可以促进发现随机池中罕见的复杂RNA基序。这是我们打算在本提案中论证和应用的主要假设。为了解决目前的限制和发展计算的体外选择，我们的目标是开发：(目标A)结构RNA池设计的计算方法；(B)筛选和测试设计的RNA池的方法；(目标C)模拟定向进化的计算方法，以优化RNA功能。在目标A中，我们将开发结构化池设计方法，允许使用蒙特卡罗模拟方法生成具有恒定结合或催化基序的用户定义的目标结构。在目标B中，我们将使用Motif扫描和筛选(例如，Pi的SVD/TNPACK软件工具)方法对设计的池的性能进行计算测试。此外，我们将通过与体外选择和纳米技术方面的RNA专家Luc Jaeger的合作，通过实验验证设计的池优于随机池。在目标C中，我们将开发一种计算方法来进行体外进化，方法包括基序扫描/筛选方法、核苷酸转移(“混合”)矩阵法和积累有益突变的偏最小二乘法，以模拟RNA基序选择和诱变PCR过程；从我们的计算体外进化方案中优化的候选RNA将由Jaeger的实验室进行实验测试。通过这些算法开发和实验合作，我们预计我们对池设计和分析以及定向进化的计算方法将提供一个全面的资源，帮助实验者设计更好的体外选择实验，并优化RNA功能，用于苛刻的生物医学应用，如发现针对癌症和其他疾病中的蛋白质的高结合亲和力适配子。我们的项目还继续为学生提供计算生物学、化学数学和生物医学方面的优秀跨学科培训。公共卫生相关性：通过体外选择和定向进化实验发现的项目叙事合成RNA具有广泛的生物和生物医学应用，包括调节疾病相关蛋白的治疗性RNA。我们的工作是基于这样的假设，即设计的结构化RNA文库比随机库更适合发现复杂的RNA。我们的项目将开发和测试用于设计结构化RNA池和增强定向进化的计算方法，以帮助实验者发现复杂的RNA。我们将与实验生物医学研究人员互动，开发和扩展我们方法的能力，以促进生物医学应用的分子工具的开发。

DESCRIPTION (provided by applicant): Project Summary/Abstract. Synthetic RNAs discovered via in vitro selection experiments have wide-ranging applications in biomedical sciences and biotechnology, including therapeutic aptamers that inhibit protein function and ribozymes that control gene expression. The RNA in vitro selection process, however, lacks systematic computational analysis to potentially increase the probability of discovering complex synthetic RNAs. To fill this significant gap, we have developed computational approaches for designing structured RNA pools and optimizing RNA functions to improve the productivity of in vitro selection and directed evolution experiments. In vitro selection experiments and our computational analysis suggest that designed RNA pools possessing diverse structural motifs can enhance discovery of complex RNA motifs which are rarely found in random pools. This is the main hypothesis we aim to demonstrate and apply in this proposal. To address current limitations and develop computational in vitro selection, we aim to develop: (Aim A) computational approaches to structured RNA pool design; (B) methods for screening and testing designed RNA pools; and (Aim C) a computational approach for simulating directed evolution for optimizing RNA functions. In Aim A, we will develop structured pool design approaches that allow generation of user-defined target structures with constant binding or catalytic motifs using a Monte Carlo simulation method. In Aim B, we will computationally test the performance of designed pools using motif scanning and screening (e.g., PI's SVD/TNPACK software tools) methods. In addition, we will experimentally verify that designed pools are superior to random pools via a collaboration with Luc Jaeger, an expert on RNA in vitro selection and nanotechnology. In Aim C, we will develop a computational approach to in vitro evolution by combining the motif scanning/screening methods, the nucleotide transition ("mixing") matrix approach, and the partial least squares method for accumulating beneficial mutations to model RNA motif selection and mutagenic PCR procedures; optimized RNA candidates from our computational in vitro evolution scheme will be experimentally tested by Jaeger's lab. With these algorithmic developments and experimental collaboration, we expect that our computational approaches to pool design and analysis and directed evolution will provide a comprehensive resource to assist experimentalists in designing better in vitro selection experiments and optimizing RNA functions for demanding biomedical applications such as discovering high-binding affinity aptamers targeting proteins in cancer and other diseases. Our project also provides continued excellent interdisciplinary training of students in computational biology, chemistry mathematics, and biomedicine. PUBLIC HEALTH RELEVANCE: Project Narrative Synthetic RNAs discovered via in vitro selection and directed evolution experiments have wide-ranging biological and biomedical applications, including therapeutic RNAs that modulate disease-related proteins. Our work is based on the hypothesis that designed structured RNA libraries are better than random pools for discovering complex RNAs. Our project will develop and test computational approaches for designing structured RNA pools and enhancing directed evolution to assist experimentalists in discovering complex RNAs. We will interact with experimental biomedical researchers to exploit and extend our methods' capabilities for advancing the development of molecular tools for biomedical applications.