An Integrated Deep Learning and Statistical Framework for Whole-Network Gene--Environment Association with Leaf Vascular Architecture

2026-07-23

AI Summary

Traditional gene–environment association studies rely on low-dimensional vein traits, overlooking the rich structural information embedded in raw leaf images. This work addresses this limitation by treating the complete leaf venation network as a high-dimensional image-based phenotype. To enable robust analysis, the authors construct high-quality annotated data by integrating EDTER and DiffusionEdge, and propose a joint modeling framework that combines semi-parametric sparse canonical correlation analysis (SSCCA) with a truncated latent Gaussian copula to handle sparse, zero-inflated edge maps. Applied to both simulated and real-world poplar datasets, the method successfully identifies three significant gene–geography interactions, demonstrating its effectiveness and generalizability in association studies involving complex image-derived phenotypes.

Abstract

Leaf veins exhibit remarkable diversity in architecture and patterning, yet existing gene--environment association studies have primarily quantified leaf venation using a small collection of low-dimensional summary traits, thereby discarding most of the structural information contained in the original images. We propose an integrated deep learning and statistical framework. The proposed framework achieves four methodological advances. First, it represents the complete leaf vascular architecture as a whole-network image phenotype. Second, it fine-tunes the deep learning-based Edge Detection with Transformers (EDTER) model to accurately extract whole-network leaf vascular architecture from RGB images by jointly learning local and global contextual features. Third, it constructs a new annotated leaf image database by integrating edge maps generated by DiffusionEdge with the Berkeley Segmentation Database (BSDS500). Fourth, it applies Semiparametric Sparse Canonical Correlation Analysis (SSCCA) to perform variable selection and model associations between repeatedly measured high-dimensional Bivariate image responses and high-dimensional predictors while simultaneously accommodating sparse, zero-inflated data represented by edge maps through a truncated latent Gaussian copula model. Two simulation studies demonstrate the performance of the proposed framework under increasing levels of complexity. Application to a real \emph{Populus} dataset identifies three significant gene--geography interactions associated with leaf vascular architecture, providing new biological insights and establishing a broadly applicable methodological framework for high-dimensional complex image phenotypes.

Problem

• gene-environment association
• leaf vascular architecture
• whole-network phenotype
• high-dimensional image data
• structural information loss

Innovation

• whole-network phenotype
• deep learning
• gene-environment interaction
• edge detection
• high-dimensional image analysis