Development of a general logistic model for disease risk prediction using multiple SNPs

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Human diseases often result from genetic variations at multiple locations, including single-nucleotide polymorphisms (SNPs).
• Accurate disease risk prediction (DRP) using SNPs is clinically significant but challenging.
• Existing commercial DRP algorithms are complex and have yielded controversial results.

Purpose of the Study:

• To develop a general, intuitive logistic model-based algorithm for disease risk prediction (DRP).
• To integrate multiple SNP risk factors from existing literature into a unified prediction model.
• To facilitate the commercialization of DRP in personalized medicine.

Main Methods:

• A logistic model was established using multiple SNP risk factors.
• SNP coefficients (β) were derived from the natural logarithm of reported odds ratios.
• A constant coefficient (β0) was determined using SNP frequencies and average population disease risk.
• Homozygous SNPs were treated as dummy variables, with SNPs subject to update.

Main Results:

• The algorithm was validated as a proof of concept.
• Two lung cancer patients were identified as maximum risk cases among 57 Chinese individuals.
• The developed DRP algorithm demonstrated intuitive and self-evident prediction capabilities.

Conclusions:

• The proposed logistic model-based DRP algorithm is more intuitive than current commercial approaches.
• This method may simplify and advance the commercialization of DRP.
• The approach holds promise for personalized medicine by leveraging genetic risk factors.