Best practices for genotype imputation from low-coverage sequencing data in natural populations

Area of Science:

• Ecological and evolutionary genetics
• Conservation biology
• Wildlife genomics

Background:

• Genetic diversity monitoring is vital for conservation, but nonmodel organisms lack specialized genotyping tools.
• High-coverage whole-genome sequencing (WGS) is accurate but expensive; reduced representation methods are cheaper but less comprehensive.
• Low-coverage WGS with imputation is a potential cost-effective alternative, but its accuracy in nonmodel populations is understudied.

Purpose of the Study:

• To empirically assess the accuracy of low-coverage WGS and imputation for genotyping in natural populations.
• To evaluate the effectiveness of this approach across different reference panel sizes (large: rhesus macaques, small: gelada monkeys).
• To provide practical recommendations for using low-coverage sequencing and imputation in wildlife genetic studies.

Main Methods:

• Sequencing of rhesus macaque (n=741) and gelada monkey (n=68) populations at low coverage (0.1-10×).
• Genotype imputation using high-confidence reference panels.
• Analysis of imputation accuracy and its impact on population genetic metrics like relatedness and population structure.

Main Results:

• High accuracy (median r² ≥ 0.92) was achieved for imputing genotypes at >95% of reference sites, even with sequencing coverage as low as 0.5×.
• Low-coverage imputed genotypes reliably estimated genetic relatedness and population structure.
• The method proved effective in both large and small reference panel populations.

Conclusions:

• Low-coverage WGS combined with imputation is an accurate and cost-effective strategy for nonmodel organism genomics.
• This approach facilitates the generation of population-scale genetic data essential for conservation biology and evolutionary research.
• Best practices and code are provided to enable wider adoption of this technique in wildlife studies.