Biomarker discovery with quantum neural networks: a case-study in CTLA4-activation pathways

Area of Science:

• Computational biology
• Quantum artificial intelligence

Background:

• Biomarker discovery from genetic data is computationally intensive due to large search spaces.
• Quantum computing and quantum AI offer potential solutions for complex biomarker discovery tasks.

Purpose of the Study:

• To propose a Quantum Neural Networks architecture for genetic biomarker discovery.
• To identify genetic biomarkers associated with specific activation pathways.

Main Methods:

• Utilized a Quantum Neural Networks architecture for biomarker identification.
• Applied Maximum Relevance-Minimum Redundancy criteria to score candidate biomarker sets.
• Demonstrated a computationally economical solution deployable on constrained hardware.

Main Results:

• Successfully demonstrated proof of concept on four CTLA4-associated activation pathways.
• Identified co-activation patterns involving CTLA4 with other genes like CD8A, CD8B, and CD2.
• Validated the model's ability to analyze complex multi-gene co-activation pathways.

Conclusions:

• The Quantum Neural Networks model identified 20 novel genetic biomarkers for CTLA4-associated pathways.
• The identified genes include CLIC4, CPE, ETS2, FAM107A, GPR116, HYOU1, LCN2, MACF1, MT1G, NAPA, NDUFS5, PAK1, PFN1, PGAP3, PPM1G, PSMD8, RNF213, SLC25A3, UBA1, and WLS.
• The implementation is open-sourced to facilitate further research in quantum AI for biomarker discovery.