Minimizing Reference Bias with an Impute-First Approach.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Pangenome indexes aim to minimize reference bias in sequencing data analysis.
• Personalized diploid references offer further bias reduction by matching individual-specific alleles.

Purpose of the Study:

• To introduce a novel impute-first alignment framework combining genotype imputation and pangenome alignment.
• To enhance the accuracy and efficiency of whole-genome DNA sequencing experiments.

Main Methods:

• Genotyping an individual using a subsample of sequencing reads.
• Imputing a personalized diploid reference using a reference panel and imputation algorithm.
• Indexing the personalized reference for read alignment with linear or graph aligners.

Main Results:

• Achieved higher variant-calling recall (99.54% vs. 99.37%) compared to graph pangenome aligners.
• Demonstrated superior precision (99.36% vs. 99.18%) and F1 scores (99.45% vs. 99.28%).
• The personalized reference index is smaller and faster to query than pangenome indexes.

Conclusions:

• The impute-first alignment framework significantly improves variant-calling performance.
• Personalized diploid references provide a more accurate and efficient alternative to pangenome indexes for DNA sequencing.
• This approach is advantageous for whole-genome DNA sequencing experiments seeking to reduce bias.