Completion of the DrugMatrix Toxicogenomics Database using 3-Dimensional Tensors

2025-07-02

AI Summary

This study addresses the high prevalence of missing data in the DrugMatrix toxicogenomics database. We propose a three-dimensional structured imputation method based on non-negative tensor completion (NTC), which explicitly models the coupled relationships among tissue type, treatment condition, and gene expression—thereby preserving the intrinsic data distribution and accurately capturing organ-specific variability. Unlike conventional matrix factorization or CP decomposition approaches, our method integrates non-negative tensor decomposition with machine learning–driven optimization. Experimental results demonstrate that it significantly outperforms existing benchmark methods in terms of mean squared error and mean absolute error. We successfully applied the method to impute the world’s largest in vivo toxicogenomics database—comprising multi-organ, multi-dose, and multi-time-point measurements—with high accuracy. This advancement substantially enhances the reliability and feasibility of cross-species drug toxicity prediction and mechanistic interpretation.

Technology Category

• Machine Learning: Matrix & Tensor Methods
• Data Mining & Knowledge Management: Data Compression
• Search and Optimization: Non-convex Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

We explore applying a tensor completion approach to complete the DrugMatrix toxicogenomics dataset. Our hypothesis is that by preserving the 3-dimensional structure of the data, which comprises tissue, treatment, and transcriptomic measurements, and by leveraging a machine learning formulation, our approach will improve upon prior state-of-the-art results. Our results demonstrate that the new tensor-based method more accurately reflects the original data distribution and effectively captures organ-specific variability. The proposed tensor-based methodology achieved lower mean squared errors and mean absolute errors compared to both conventional Canonical Polyadic decomposition and 2-dimensional matrix factorization methods. In addition, our non-negative tensor completion implementation reveals relationships among tissues. Our findings not only complete the world's largest in-vivo toxicogenomics database with improved accuracy but also offer a promising methodology for future studies of drugs that may cross species barriers, for example, from rats to humans.

Problem

• Complete DrugMatrix toxicogenomics data using tensor completion
• Improve accuracy by preserving 3D data structure and machine learning
• Capture organ-specific variability and cross-species drug relationships

Innovation

• Tensor completion preserves 3D data structure
• Non-negative tensor reveals tissue relationships
• Lower errors than matrix factorization methods