An official website of the United States government.

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

APPLYING MACHINE LEARNING TO MICROBIOME ANALYSIS AS A TOOL TO IMPROVE POULTRY PRODUCTION.

Objective

The goal of this project is to produce an ML model that can provide predictive and diagnostic information to poultry production managers, and to train a new generation of computer specialists skilled in ML/AI as applied to animal science.Objective 1 - Conduct experimental pen trials using probiotics to determine how treatments affect the bacteriomes and key indicators of growth, pathogen presence, and immune status. Drs. Barnas and Timmons will oversee live production and sample collection, Dr. Taabodi will perform analytical procedures for immune response and pathogen prevalence and Dr. May will coordinate all activities including sample distribution for analyses. Shotgun analyses will be performed by a commercial operation working with USDA-ARS.Objective 2 - Develop a meta-analysis tool for machine learning of 16S / shotgun metagenomics using data from objective 1. Drs. Barnas, Davies, May, and Summers will be involved in collecting data from published papers and Drs. Davies and Summers will develop the meta-analysis tools. Dr. Green-Miller (letters of collaboration and CV attached from the AI Farms National AI Institute)Objective 3 - Collect feed, litter, and cecal samples from the same four commercial chicken houses over three seasons (Spring, Summer, Fall) to determine how poultry house conditions affect the bacteriomes and key indicators of growth, pathogen presence, and immune status of broilers. Dr. Taabodi will perform analytical procedures for immune response and pathogen prevalence and Dr. Timmons will evaluate the growth performance and Dr. May will coordinate all activities including sample distribution for analyses. Shotgun analyses will be performed by a commercial operation working with USDA-ARS.Objective 4 - Input all data into machine learning models to further refine the program such that it can predict outcome (i.e. growth performance). Drs. Davies and Summers will be responsible for this portion in conjunction with the student interns. Drs. Barnas and Timmons will communicate results to growers and the poultry industry.Objective 5 - Provide educational opportunities for 10 undergraduate students as interns to the USDA-ARS Beltsville and additional tuition remission for an additional 10 undergraduate students. Support 2 graduate students (1 Ph.D. and 1 M.S.). All of the personnel listed will be involved in this part of the project by serving as members of the graduate committees.

Investigators
May, E. B.; Timmons, JE, .; Summers, KA, LY.; Barnas, MI, R.; Taabodi, MA, .; Pirone-Davies, CA, .
Institution
UNIVERSITY OF MARYLAND EASTERN SHORE
Start date
2022
End date
2025
Project number
MDX-ASEQ20220501
Accession number
1028568
Commodities