GenePheno: Interpretable Gene Knockout-Induced Phenotype Abnormality Prediction from Gene Sequences

2025-11-12

AI Summary

Existing methods struggle to generalize from genomic sequences to predict multi-phenotypic abnormalities induced by gene knockouts, hindered by modality gaps, pleiotropy, and reliance on manual annotations. This paper introduces the first end-to-end interpretable framework that directly maps raw gene sequences to multi-phenotype predictions. It integrates contrastive multi-label learning, exclusive regularization, and a gene-functional bottleneck layer to jointly predict phenotypes while uncovering underlying biological mechanisms. Evaluated on four in-house datasets, the model achieves significant improvements in gene-centric F_max and phenotype-centric AUC. Case studies demonstrate its ability to accurately recover known gene functional pathways. By bridging the sequence-to-phenotype mapping gap, this work establishes a novel paradigm for high-throughput, prior-free gene function interpretation.

Technology Category

• Machine Learning: Multimodal Learning
• Game Theory and Economic Paradigms: Adversarial Learning
• Computer Vision: Multi-modal Vision

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Exploring how genetic sequences shape phenotypes is a fundamental challenge in biology and a key step toward scalable, hypothesis-driven experimentation. The task is complicated by the large modality gap between sequences and phenotypes, as well as the pleiotropic nature of gene-phenotype relationships. Existing sequence-based efforts focus on the degree to which variants of specific genes alter a limited set of phenotypes, while general gene knockout induced phenotype abnormality prediction methods heavily rely on curated genetic information as inputs, which limits scalability and generalizability. As a result, the task of broadly predicting the presence of multiple phenotype abnormalities under gene knockout directly from gene sequences remains underexplored. We introduce GenePheno, the first interpretable multi-label prediction framework that predicts knockout induced phenotypic abnormalities from gene sequences. GenePheno employs a contrastive multi-label learning objective that captures inter-phenotype correlations, complemented by an exclusive regularization that enforces biological consistency. It further incorporates a gene function bottleneck layer, offering human interpretable concepts that reflect functional mechanisms behind phenotype formation. To support progress in this area, we curate four datasets with canonical gene sequences as input and multi-label phenotypic abnormalities induced by gene knockouts as targets. Across these datasets, GenePheno achieves state-of-the-art gene-centric Fmax and phenotype-centric AUC, and case studies demonstrate its ability to reveal gene functional mechanisms.

Problem

• Predicts gene knockout-induced phenotype abnormalities from genetic sequences
• Addresses modality gap between gene sequences and pleiotropic phenotypes
• Provides interpretable multi-label predictions with biological consistency constraints

Innovation

• Interpretable multi-label prediction from gene sequences
• Contrastive learning with exclusive regularization for correlations
• Gene function bottleneck layer for biological interpretability