MultiCook: A Tool That Improves Accuracy of HLA Imputation by Combining Probabilities From Multiple Reference Panels...

Area of Science:

• Immunogenetics
• Computational Biology
• Genomics

Background:

• Human Leukocyte Antigen (HLA) genotype imputation is crucial for genetic studies but traditionally costly.
• Current computational methods using single nucleotide polymorphism (SNP) markers offer a cost-effective alternative.
• Existing imputation methods require ethnicity-matched reference panels, which are often limited in size and availability.

Purpose of the Study:

• To introduce MultiCook, a novel computational tool for enhancing HLA genotype imputation.
• To enable the simultaneous utilization of multiple HLA imputation methods and reference panels.
• To improve imputation accuracy and efficiency, particularly for diverse populations.

Main Methods:

• Developed MultiCook to integrate multiple HLA imputation algorithms and reference panels.
• Enabled merging of panels from different SNP genotyping platforms.
• Allowed combination of outputs from various imputation methods and the Michigan Imputation Server.

Main Results:

• MultiCook reduced the imputation error rate by approximately one third in a Korean cohort (from 4.70% to 3.31%) by combining East Asian and Han Chinese reference panels.
• Demonstrated improved accuracy for imputing low-frequency HLA alleles compared to single-panel approaches and the Michigan HLA imputation server.
• Showcased versatility in merging diverse reference panels and imputation method outputs.

Conclusions:

• MultiCook offers a versatile and effective solution for improving HLA genotype imputation accuracy.
• The tool addresses the limitations of scarce, ethnicity-specific reference panels by enabling the combined use of multiple resources.
• MultiCook enhances the accessibility and reliability of HLA imputation for a wider range of populations and research applications.