Towards a gamete matching platform: using immunogenetics and artificial intelligence to predict recurrent...

Area of Science:

• Immunogenetics
• Reproductive Medicine
• Computational Biology

Background:

• Recurrent miscarriage (RM) is associated with Human Leukocyte Antigen (HLA) allele sharing between partners.
• The complex genetic interactions between parents hinder the identification of clear genetic markers for RM.
• Existing methods lack the precision to fully elucidate the genetic underpinnings of RM.

Purpose of the Study:

• To develop a novel methodology for analyzing HLA haplotypes in couples with and without a history of RM.
• To create an algorithm for calculating genetic risk associated with recurrent miscarriage.
• To explore the potential of combining immunogenetics with artificial intelligence for improved reproductive outcomes.

Main Methods:

• High-resolution HLA typing of parental haplotypes.
• Application of linear algebra on peptide binding affinity data.
• Development of the IMMATCH algorithm for retrospective RM prediction.

Main Results:

• The IMMATCH algorithm achieved an Area Under the Curve (AUC) of 0.71 (p=0.0035) in predicting RM.
• The algorithm demonstrated a sensitivity of 86% with an adjustable threshold.
• The study highlights the efficacy of integrating immunogenetic data with AI for risk assessment.

Conclusions:

• The proposed methodology offers a promising approach to identify genetic risk factors for recurrent miscarriage.
• Combining immunogenetics and AI can lead to personalized tools for understanding unexplained infertility.
• This research paves the way for a gamete matching platform to enhance pregnancy success rates.