Pan-African genome demonstrates how population-specific genome graphs improve high-throughput sequencing data...

Area of Science:

• Bioinformatics
• Genomics
• Computational Biology

Background:

• The linear human genome reference struggles to represent population diversity and maintain accuracy for non-European ancestries.
• Existing graph-based genomics toolkits focus on read alignment and variant calling, neglecting variant curation and graph construction methods.

Purpose of the Study:

• To address challenges in constructing accurate genome graphs.
• To propose methods for sample selection, graph augmentation, and resolving reference ambiguity.
• To demonstrate the effectiveness of population-specific genome graphs.

Main Methods:

• Developed methods for sample selection based on population diversity.
• Proposed graph augmentation strategies incorporating structural variants.
• Introduced techniques to resolve graph reference ambiguity.
• Iteratively augmented tailored genome graphs using whole-genome samples of African ancestry.

Main Results:

• Population-specific genome graphs significantly reduce read mapping errors compared to linear or generic graphs.
• Enhanced variant calling sensitivity was achieved with tailored graph references.
• Joint variant calling improvements were realized without intensive post-processing.

Conclusions:

• Population-specific genome graphs are more representative than linear or generic references.
• Tailored genome graphs offer a more accurate and efficient approach for genomic analysis in diverse populations.
• This approach improves the overall bioinformatics pipeline for variant discovery.