BMC Bioinformatics, Год журнала: 2024, Номер 25(1)

Опубликована: Ноя. 26, 2024

Abstract Background The preservation of soil health is a critical challenge in the 21st century due to its significant impact on agriculture, human health, and biodiversity. We provide one first comprehensive investigations into predictive potential machine learning models for understanding connections between biological phenotypes. investigate an integrative framework performing accurate learning-based prediction plant performance from biological, chemical, physical properties via two models: random forest Bayesian neural network. Results Prediction improves when we add environmental features, such as microbial density, along with microbiome data. Different preprocessing strategies show that decisions significantly performance. naive total sum scaling normalization commonly used research optimal maximize power. Also, find accurately defined labels are more important than normalization, taxonomic level, or model characteristics. ML limited humans can’t classify samples accurately. Lastly, domain scientists full selection decision tree identify choices optimize Conclusions Our study highlights importance incorporating diverse features careful data enhancing power phenotype connections. This approach can contribute advancing agricultural practices management.

Язык: Английский