Using multiple sampling strategies to estimate SARS-CoV-2 epidemiological parameters from genomic sequencing data.

Area of Science:

• Genomics
• Epidemiology
• Bioinformatics

Background:

• Viral genomic data offers rich insights for epidemiological analysis.
• Sequence selection can introduce biases, potentially diminishing the value of these datasets.
• The impact of different sampling strategies on genomic analysis is understudied.

Purpose of the Study:

• To investigate the influence of viral sequence selection on epidemiological parameter estimation.
• To evaluate how various sampling schemes affect estimates of transmission rates and origin.
• To provide insights into optimal sequence selection for SARS-CoV-2 genomic analysis.

Main Methods:

• Utilized SARS-CoV-2 genomic sequences from Hong Kong and Amazonas State, Brazil.
• Applied multiple sampling schemes to estimate key epidemiological parameters.
• Compared estimates derived from different sampling strategies, including unsampled datasets.

Main Results:

• Real-time (R_t) and per-capita (r_t) transmission rates are sensitive to sampling variations.
• Basic reproduction number (R₀) and date of origin estimates showed relative robustness to sampling changes.
• Analysis of unsampled datasets yielded the most biased R_t and r_t estimates in both case studies.

Conclusions:

• Sampling strategy is a critical, yet often overlooked, factor in genomic sequencing analysis pipelines.
• Biases introduced by sequence selection can significantly impact epidemiological conclusions.
• Careful consideration of sampling methods is essential for accurate viral genetic and epidemiological analyses.