文章基本信息

标题：Evaluation of Linear Models and Linear Mixed Models to Predict the Effects of Antimicrobial Peptides on Broiler Performance
本地全文：下载
作者：M M Sholikin ; M D Alifian ; A T Wahyudi 等
期刊名称：IOP Conference Series: Earth and Environmental Science
印刷版ISSN：1755-1307
电子版ISSN：1755-1315
出版年度：2020
卷号：478
期号：1
DOI：10.1088/1755-1315/478/1/012002
语种：English
出版社：IOP Publishing
摘要：This study aims to evaluate linear models and linear mixed models that were built based on the database of effect Antimicrobial Peptide (AMP) on broiler performance. The method of this research consists of compiling a database, making models, and evaluating models. The database compiled from a comparative study related to the effect of AMP on body weight, average daily gain, daily feed intake, and the feed conversion ratio from broilers reared at 1-21 days. The model consists of five Linear Models (LM) and five Linear Mixed Models (LMM). The best model determined by Bayesian Information Criteria (BIC) and then the correlation value calculated according to actual vs. predicted value. Based on the lowest BIC value, the representative model for body weight, average body weight gain, and average feed consumption are model 9. Besides that, the feed conversion ratio relates to model 2. Model 9 is a linear mixed model with levels as fixed effects and random effects consisting of studies, broiler breeds, type of peptides, and treatment methods. In conclusion, LMM is better than LM in predicting complex problems such as comparative studies of the effects AMP on broiler performance.