Flexynesis: A deep learning toolkit for bulk multi-omics data integration for precision oncology and beyond
Bora Uyar, Taras Savchyn, Amirhossein Naghsh Nilchi, Ahmet Sarıgün, Ricardo Wurmus, Mohammed Maqsood Shaik, Björn Grüning, Vedran Franke, Altuna Akalin
Abstract
Accurate decision making in precision oncology depends on integration of multimodal molecular information, for which various deep learning methods have been developed. However, most deep learning-based bulk multi-omics integration methods lack transparency, modularity, deployability, and are limited to narrow tasks. To address these limitations, we introduce Flexynesis, which streamlines data processing, feature selection, hyperparameter tuning, and marker discovery. Users can choose from deep learning architectures or classical supervised machine learning methods with a standardized input interface for single/multi-task training and evaluation for regression, classification, and survival modeling. We showcase the tool's capability across diverse use-cases in precision oncology. To maximize accessibility, Flexynesis is available on PyPi, Guix, Bioconda, and the Galaxy Server ( https://usegalaxy.eu/ ). This toolset makes deep-learning based bulk multi-omics data integration in clinical/pre-clinical research more accessible to users with or without deep-learning experience. Flexynesis is available at https://github.com/BIMSBbioinfo/flexynesis .