A framework for intelligent data acquisition and real-time database searching for shotgun proteomics

Area of Science:

• Proteomics
• Computational Biology
• Mass Spectrometry

Background:

• Complex peptide mixtures in mass spectrometry-based proteomics overwhelm instrument capacity.
• Optimal allocation of peptide sequencing and quantification events is crucial for efficient analysis.
• Intelligent agents offer autonomous decision-making capabilities in complex environments.

Purpose of the Study:

• To develop and describe a framework for intelligent data acquisition and real-time database searching in proteomics.
• To implement an intelligent agent within the MaxQuant computational environment (MaxQuant Real-Time).
• To showcase the capabilities of this framework in enhancing peptide identification and quantification.

Main Methods:

• Development of an intelligent agent integrated into MaxQuant.
• Real-time analysis of mass spectrometry data as it is acquired.
• Construction of isotope patterns and SILAC pair information.
• Control of MS and tandem MS events based on real-time and prior data.
• Implementation of a real-time search engine for on-the-fly peptide identification.

Main Results:

• The intelligent agent achieves performance comparable to existing data-dependent methods.
• Peptide identification is achieved within 30 ms using the real-time search engine.
• Sequencing events can be focused on peptides of specific interest (e.g., by Gene Ontology term or modifications).
• Enhanced quantification of SILAC pairs with previously poorly defined ratios.

Conclusions:

• MaxQuant Real-Time provides a flexible framework for intelligent, directed data acquisition in proteomics.
• This approach is particularly beneficial for new instrument types like the quadrupole-Orbitrap.
• The system enables more efficient and targeted analysis of complex peptide mixtures.