AMFinder

AMFinder identifies and quantifies arbuscular mycorrhizal (AM) fungal colonization and intraradical hyphal structures in ink-stained plant root images using convolutional neural networks (CNNs).

Key Features:

• Automatic identification and quantification: Automates detection and measurement of AM fungal colonization on ink-stained root samples.
• Convolutional neural networks (CNNs): Uses CNNs to interpret complex patterns within ink-stained images for identification of mycorrhizal structures.
• Versatility across species and conditions: Validated on Nicotiana benthamiana, Medicago truncatula, Lotus japonicus, and Oryza sativa and applicable across varied experimental conditions.
• Support for mutant analysis: Detects altered colonization patterns in mutant strains such as ram1-1 and strandsmax1.
• Dynamic colonization tracking: Quantifies temporal changes in fungal colonization across whole root systems.

Scientific Applications:

• Analysis of plant–fungus mutualisms: Enables quantitative study of arbuscular mycorrhizal fungal interactions with plant roots.
• Ecological and temporal studies: Supports investigation of dynamic colonization patterns and their implications for plant health and soil ecology.
• Genetic analysis of mutants: Provides quantitative phenotyping of colonization in mutant plants, including ram1-1 and strandsmax1.

Methodology:

Image acquisition of ink-stained roots by flatbed scanning or digital microscopy followed by analysis using convolutional neural networks (CNNs).