Simulated SARS-CoV-2 Wastewater Sequencing Data for Variant Abundance Benchmarking

A collection of benchmarking datasets simulating wastewater sequencing reads for SARS-CoV-2 variants B.1.1.7, B.1.351, B.1.427, B.1.429, and P.1. Variant abundance ranges from 0.05% to 100% across 33 benchmarks per variant, with simulations for whole genomes and Spike-only sequences at coverages of 100x, 1000x, and 10,000x. The data was created by Jasmijn A. Baaijens of Harvard University to evaluate the accuracy of variant abundance predictions from wastewater sequencing.

Use Cases

• Benchmarking variant abundance estimation algorithms based on simulated whole genome sequencing reads.
• Evaluating the impact of sequencing depth on detection sensitivity using the 100x, 1000x, and 10,000x coverage benchmarks.
• Comparing variant detection performance between whole genome and Spike-only (S-only) sequencing strategies.
• Assessing the lower detection limit for SARS-CoV-2 variants in wastewater based on the 0.05% to 100% abundance range.

Strengths

• Simulated data for five key SARS-CoV-2 variants (B.1.1.7, B.1.351, B.1.427, B.1.429, P.1) provides a controlled test environment.
• Systematic variant abundance gradient from 0.05% to 100% across 33 benchmarks allows for sensitivity analysis.
• Includes simulations at multiple sequencing depths (100x, 1000x, 10,000x) to evaluate coverage effects.
• Provides both whole genome and Spike-gene-only (S-only) benchmark sets, reflecting different sequencing approaches.

Limitations

• Data is simulated and may not capture all complexities of real wastewater sequencing samples.
• Row count and file size are unknown, which may limit suitability assessment for large-scale processing.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

Jasmijn A. Baaijens, Harvard University, via paperswithcode.

Collection Method

Computational simulation of sequencing reads from variant and background SARS-CoV-2 genomes.