CoDON: a learning framework for linking genomics and transcriptomics data to protein expression

Motivation

• Explain the protein expression profiles through the integration of genomic and transcriptomic data
• Link different data types in the context of complex diseases
• Wide range of applications (e.g.: multi-omics perturbation analysis)

Aims

• Reconstruct proteins profiles from different data layers using a common learning framework
• Understand how different omic datasets can be combined in a common reduced representation
• Decipher highly non-linear molecular interactions