LEAP-HI/GOALI: Engineering Crops for Genetic Adaptation to Changing Enviroments

这个美国繁荣、健康和基础设施领先工程（LEAP-HI）学术与工业联络机会（GOALI）项目解决了NSF理解生命规则和利用数据革命的大思路，目标是在不确定和快速变化的环境中使用更少的资源（土地、水、农药和化肥）为不断增长的人口提供食物、纤维和燃料的需求。人们普遍认识到，目前的农业技术，从作物遗传改良到大田作物生产，将无法满足未来的农业需求，因为它们严重依赖昂贵、耗时、试验和错误的田间试验来开发改良的植物品种。用于分析高维数据的新兴数学优化和机器学习方法为加快植物育种提供了机会，以实现作物对不断变化的环境的快速有效适应。本项目的方法将利用工程技术，这些技术已被用于显著提高通信、制造、运输和能源系统的效率和弹性。该研究需要综合多个学科，包括农学、作物建模、机器学习、运筹学、优化和植物育种，旨在展示工程在解决农业挑战方面的领导作用。将解决三个技术问题，它们代表了农艺系统的一个小但高度可见的子集：(1)基于遗传、农艺管理和环境数据及其相互作用准确预测植物表型；(2)设计遗传改良体系，高效培育表型优良品种；(3)设计作物管理策略，确保作物在不断变化的环境下获得优越的表型，同时在决策过程中平衡回报、时间和风险。研究团队将首先将技术问题转化为工程目标，然后确定现有方法并设计新的方法来实现目标。相应的工程目标是：(1)确定与协同效应相关的一小部分变量，以及它们的加性效应；(2)设计了一套基因组选择算法，该算法是一种特殊类型的非线性、非凸、高维、受资源可用性和生殖生物学规律约束的动态优化问题；(3)建立一套多目标、多层次的优化模型和算法，以平衡受遗传、环境和后勤约束的回报、时间和风险。实现这些目标将证明工程方法在提高农艺系统的效率和弹性方面的力量，其目的是将植物育种建立为一门工程学科。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This Leading Engineering for America's Prosperity, Health, and Infrastructure (LEAP-HI) Grant Opportunities for Academic Liaison with Industry (GOALI) project addresses the NSF Big Ideas of Understanding the Rules of Life and Harnessing the Data Revolution in targeting the need to provide food, fiber and fuel for a growing population using fewer resources (land, water, pesticides and fertilizers) in uncertain and rapidly changing environments. It is widely recognized that current agricultural technologies, from crop genetic improvement to field crop production, will not meet future agricultural demands, due to their heavy reliance on expensive, time-consuming, trail and error field trials to develop improved plant breeds. Emerging mathematical optimization and machine learning methods for analyzing high-dimensional data provide opportunities to speed up plant breeding to achieve rapid and efficient adaptation of crops to changing environments. The approaches in this project will take advantage of engineering techniques that have been used to remarkably improve the efficiency and resiliency of communication, manufacturing, transportation and energy systems. The research requires the synthesis of multiple disciplines, including agronomy, crop modeling, machine learning, operations research, optimization and plant breeding and aims to demonstrate the leadership role of engineering in addressing agricultural challenges.Three technical issues, which represent a small but highly visible subset of agronomic systems, will be addressed: (1) accurately predicting plant phenotypes based on genetic, agronomic management and environmental data and their interactions; (2) design of genetic improvement systems to efficiently develop cultivars with superior phenotypes; and (3) design of crop management strategies to assure that crops achieve superior phenotypes under changing environments, while balancing reward, time, and risk in the decision-making process. The research team will first translate the technical issues into engineering objectives and then identify existing methods and design new ones to achieve the objectives. The corresponding engineering objectives are: (1) identify a small subset of variables associated with synergistic effects in addition to their additive effects; (2) design a set of algorithms for genomic selection, which is a special type of nonlinear, non-convex, high-dimensional, and dynamic optimization problem constrained by resource availability and laws of reproductive biology; and (3) create a set of multi-objective and multi-level optimization models and algorithms for balancing reward, time, and risk, subject to genetic, environmental, and logistical constraints. Achieving these objectives will demonstrate the power of engineering approaches in improving the efficiency and resiliency of agronomic systems, with the aim of establishing plant breeding as an engineering discipline.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.