PanEffect

PanEffect computes and visualizes variant effect scores across the maize pan-genome to quantify potential functional impacts of coding variants.

Key Features:

• Comprehensive Variant Analysis: Analyzes over 550 million potential amino acid substitutions within the B73 maize reference genome and approximately 2.3 million natural variations across the maize pan-genome.
• Advanced Scoring System: Uses the Evolutionary Scale Modeling (ESM) protein language model to calculate variant effect scores from log-likelihood ratio differences between the B73 reference sequence and pan-genome variants.
• Visualization Capabilities: Represents variant effect scores as heatmaps spanning a spectrum from benign outcomes to significant functional consequences.
• Functional and Structural Context: Provides per-variant displays of secondary structures and annotated functional domains to contextualize predicted impacts on protein structure and function.

Scientific Applications:

• Pan-genome variant exploration: Enable large-scale exploration of protein-coding variants across multiple maize genomes.
• Trait and functional outcome prediction: Support prioritization of variants for predicting traits and functional consequences based on effect scores.
• Genetic target identification for crop improvement: Facilitate identification of candidate high-impact variants for use in crop enhancement and breeding studies.

Methodology:

Variant effect scores are computed with the ESM protein language model using log-likelihood ratio differences between the B73 reference and pan-genome variants; outputs include heatmaps alongside secondary structure and functional domain annotations for coding variants across maize genomes.