Performing Department
(N/A)
Non Technical Summary
The FARAD program is a national food safety program funded for more than 40 years since 1982 by USDA. The program a collaborative effort with five regional centers: Kansas State University--Olathe (KSUO), North Carolina State University, University of California--Davis, University of Florida, and Virginia-Maryland College of Veterinary Medicine. The goal of FARAD is to provide the most updated information and scientific tools to help the production of safe foods of animal origin. The program accomplishes this goal through its objectives: (1) to identify, extract, assemble, evaluate and distribute reviewed information about residue avoidance and mitigation to people involved in residue avoidance programs; (2) to develop tools that allow people to predict proper withdrawal intervals after extralabel drug use. To continue to fulfill the mission of FARAD, during 2022-2023, the specific objectives at KSUO include: (1) to develop interfaces by which FARAD responders can access the KSUO comparative medicine databases and associated computational tools, as well as accessing relevant global data, (2) to pursue the goal of developing novel AI technologies practical to food-animal production for collecting, organizing, merging, and cleaning data and help easily submit queries to obtain information regarding the datasets of Global maximum residue limits (MRLs) and withdrawal periods (WDPs), and (3) to conduct research in designing a full-text retrieval system that will enable the creation of an information retrieval database that can be easily queried, performing metadata analysis on bibliographic records in veterinary medicine as well as those validated by the current FARAD database.
Animal Health Component
100%
Research Effort Categories
Basic
60%
Applied
20%
Developmental
20%
Goals / Objectives
The goal of the national FARAD program is to protect the American public by promoting the production of safe, animal-derived human food products (milk, eggs, meat, honey, etc.) that are devoid of violative or potentially unsafe chemical residues, including drugs, pesticides, environmental contaminants, natural toxins, and other harmful substances. Kansas State University-Olathe (KSUO) will be responsible for developing interfaces by which FARAD responders can access the KSUO comparative medicine databases and associated computational tools, as well as access relevant global data. The KSUO team will pursue the goal of developing novel AI technologies practical to food-animal production for collecting, organizing, merging, and cleaning data and help easily submit queries to obtain information regarding the datasets of Global maximum residue limits (MRLs) and withdrawal periods (WDPs). They will also conduct research in designing a full-text retrieval system that will enable the creation of an information retrieval database that can be easily queried, performing metadata analysis on bibliographic records in veterinary medicine as well as those validated by the current FARAD database. This information retrieval system is significant and will be used to develop hybrid models of content-based and bibliometric features for machine learning applications to extract and clean data from different resources as an integral part of the next-generation FARAD program.
Project Methods
The methods employed in this project involve comprehensive training of many students in data analytics techniques, particularly focusing on database curation, decision-making, and result interpretation. Students will be equipped with the skills to develop working hypotheses and design error correction techniques for data scrubbing and retrieval. They will proceed with data exploration techniques to visually analyze and comprehend the data's characteristics. Subsequently, data curation and annotation will be carried out to organize and integrate data from diverse sources, involving annotation, organization, clustering, and presentation of data types from the 1DATA databank.Integration of machine learning models will follow, enabling the acquisition of results after data preprocessing and cleansing, effectively reducing data size and eliminating insignificant and noise-driven reports. This process will ultimately enhance decision-making and interpretation through data-driven machine learning. Notably, statistical methods such as regression analysis, cohort and cluster analysis, feature reduction, and data visualization will be utilized extensively throughout the project for data analytics.