Hap-E Search 2.0: Improving the Performance of a Probabilistic Donor-Recipient Matching Algorithm Based on Haplotype...

Area of Science:

• Immunogenetics
• Transplantation immunology
• Computational biology

Background:

• Human leukocyte antigen (HLA) matching is critical for hematopoietic stem cell transplantation (HSCT) donor selection.
• Current HLA matching algorithms are computationally intensive, leading to lengthy search times.
• Ambiguous or incomplete HLA typing data complicates donor identification.

Purpose of the Study:

• To introduce a novel, optimized search kernel for the Hap-E Search algorithm.
• To significantly reduce the computational time required for HLA donor matching in large databases.
• To validate the accuracy of the new algorithm's matching probability predictions.

Main Methods:

• Implementation of a new search kernel utilizing pre-calculated donor genotype information.
• Testing the updated Hap-E Search algorithm on a database exceeding 9 million donors.
• Validation of predicted matching probabilities using 67,550 donor-patient pair verification typing requests.

Main Results:

• The new search kernel provides HLA matching results in under one minute, even for large databases.
• Search acceleration factors of 10 or more were observed compared to the previous implementation.
• The algorithm efficiently handles frequent or incomplete patient HLA data, yielding thousands of potential donors.

Conclusions:

• The enhanced Hap-E Search algorithm offers a substantial improvement in speed and efficiency for HLA donor matching.
• This advancement facilitates quicker identification of suitable donors for hematopoietic stem cell transplantation.
• The validated accuracy ensures reliable donor selection for improved patient outcomes.