Excellence in Research: Aptamer integrated graphene-gold conjugates for machine learning aided pesticide residue screening

本论文的主要目的是设计石墨烯-金-适体生物复合物用于构建比色农药传感器。所设计的生物缀合物将被定制用于在柔性基底上的喷墨印刷。比色输出将由相机集成手持设备或智能手机相机捕获。根据捕获设备和照明条件，将使用机器学习算法进行校正，以提取与农药水平相关的精确色度信息。为了证明所提出的平台的通用性，四种农药，即草甘膦，马拉硫磷，啶虫脒，毒死蜱检测的案例研究将进行。作为该项目的一部分，诺福克州立大学工程系的研究生和本科生将通过应用工程，纳米技术，图像处理和机器学习的概念在设计传感系统方面获得宝贵的多学科经验。该项目还将通过提供课程机会，如独立学习，暑期实习和高级顶点项目来加强。项目展示和示范将包括在工程部推广到诺福克地区的高中学生和教师。所提出的生物缀合物是基于石墨烯-金缀合物的过氧化物酶样纳米酶活性、适体的选择性和机器学习的可靠性。所提出的架构可以作为一个通用的比色传感平台，通过改变不同目标的适体序列，提供了一个强大的和通用的工程系统，这是可扩展的，允许测试范围广泛的农药。研究任务的成果不仅将为农药分析提供快速分析工具，而且还将产生关于适体-小分子靶标结合的前所未有的新知识。为了准确地将色度变化与农药水平相关联，将使用机器学习方法将照度元素与色度因子分离。使用RGB，LAB和HSV颜色空间的传感器的色度输出上的不同的照明条件的影响将被调查。机器学习方法还将提供理解实验参数（如样品pH值和电导率、适体长度和%GC）对适体-靶标亲和力的影响所需的方法学框架。开发的平台将能够测量结合动力学，监测适体的平衡亲和力，并调整非特异性相互作用的存在。此外，由于传感器制造基于喷墨印刷和热压印等低成本技术，因此该技术具有很高的商业潜力。该奖项反映了NSF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。

The main objective of this proposal is to design graphene-gold-aptamer bioconjugates for building colorimetric pesticide sensors. The bioconjugates designed will be tailored for inkjet printing on flexible substrates. The colorimetric output will be captured by a camera integrated handheld device or a smartphone camera. Dependency on capture device and illumination conditions will be corrected using a machine learning algorithm to extract the precise chromatic information that correlate to pesticide levels. In order to prove the versatility of the proposed platform, a case study on detection of four pesticides namely Glyphosate, Malathion, Acetamiprid, and Chlorpyrifos will be performed. By being part of this project, graduate and undergraduate students of the Department of Engineering, Norfolk State University, will gain valuable multidisciplinary experience in designing sensing systems by applying concepts of engineering, nanotechnology, image processing and machine learning. This project will also be enhanced by offering curricular opportunities such as independent study, summer internships, and senior capstone projects. Project displays and demonstrations will be included in Engineering Department outreach to Norfolk area high school students and teachers. The proposed bioconjugate is based on the peroxidase like nanozyme activity of the Graphene-gold conjugate, selectivity of the aptamer and reliability of machine learning. The proposed architecture can serve as a universal colorimetric sensing platform by changing the aptamer sequence for different targets, providing a powerful and versatile engineered system which is scalable, allowing testing of a wide range of pesticides. Products of the research tasks will not only result in a rapid analysis tool for pesticide analysis but will also yield unprecedented new knowledge on aptamer-small molecule target binding. In order to accurately correlate chrominance changes to pesticide levels, the illuminance element will be separated from the chrominance factor using a machine learning approach. The effect of varying illumination conditions on the colorimetric output of the sensor using RGB, LAB and HSV color spaces will be investigated. The machine learning approach will also provide the methodological framework needed to understand the influence of experimental parameters such as sample pH and conductivity, aptamer length, and %GC on aptamer-target affinity. The platform developed will be capable of measuring binding kinetics, monitoring equilibrium affinities of aptamers and adjusting for the presence of nonspecific interactions. Moreover, this technology has high commercial potential as the sensor fabrication is based on low-cost techniques like inkjet printing and thermal embossing.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.