G2PDeep-v2: a web-based deep-learning framework for phenotype prediction and biomarker discovery using multi-omics...

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Phenotype prediction and genetic marker discovery are crucial for understanding complex biological processes and diseases.
• Multi-omics data integration presents challenges for traditional analytical methods.

Purpose of the Study:

• To introduce G2PDeep-v2, a novel deep learning-based web server for phenotype prediction and marker discovery.
• To provide a user-friendly platform for researchers to analyze multi-omics data across various organisms.

Main Methods:

• Development of a deep learning framework (G2PDeep-v2) accessible via a web server.
• Implementation of an automated hyperparameter tuning algorithm for model training on high-performance computing resources.
• Integration of visualization tools for prediction results and Gene Set Enrichment Analysis (GSEA).

Main Results:

• G2PDeep-v2 enables accurate phenotype prediction and identification of significant genetic markers from multi-omics data.
• The platform supports diverse organisms, including humans, plants, animals, and viruses.
• Users can gain insights into molecular mechanisms through interactive analysis and GSEA.

Conclusions:

• G2PDeep-v2 offers a powerful and accessible tool for advancing biological research through deep learning.
• The platform facilitates the discovery of genotype-phenotype relationships and underlying molecular pathways.
• G2PDeep-v2 democratizes advanced multi-omics data analysis for a wide range of scientific applications.