Machine learning applications in proteomics research: how the past can boost the future

Area of Science:

• Proteomics
• Artificial Intelligence
• Computational Biology

Background:

• Mass spectrometry (MS)-based proteomics involves complex problems.
• Increasing availability of public data fuels computational approaches.
• Machine learning (ML) is emerging as a key tool in proteomics research.

Purpose of the Study:

• To provide a comprehensive overview of ML applications in proteomics.
• To cover the entire proteomics workflow, including wet-lab and dry-lab processes.
• To highlight how ML addresses critical bottlenecks in experiment planning, design, data processing, and analysis.

Main Methods:

• Review of existing literature on ML applications in proteomics.
• Categorization of ML applications across the proteomics workflow.
• Identification of key challenges and bottlenecks addressed by ML.

Main Results:

• ML applications span experiment planning, design, and data processing.
• ML algorithms are used to solve intractable problems in proteomics.
• The overview covers diverse areas from sample preparation to data interpretation.

Conclusions:

• Machine learning is a rapidly growing and valuable tool in MS-based proteomics.
• ML effectively addresses significant bottlenecks throughout the proteomics experimental and analytical workflow.
• The integration of ML enhances the efficiency and scope of proteomics research.