25 Synthetic TP53 Genomic Datasets for Benchmarking and Method Development

This collection contains 25 synthetic genomics datasets generated using NEAT v3, simulating the TP53 gene of Homo sapiens. These datasets are intended for benchmarking somatic variant calling algorithms, especially in tumor-only settings.

Each dataset is composed of paired-end reads and was designed to systematically explore the impact of two intrinsic NGS parameters on variant detection performance:

Sequencing coverage: 300×, 700×, 1000×, 3000×, and 5000×

Read length: 50 bp, 75 bp, 100 bp, 150 bp, and 300 bp

All 25 datasets are labeled using the format {coverage}_{readLength} to reflect their specific configuration. For example, the dataset 1000_150 corresponds to 1000× coverage and 150 bp read length. All combinations are shown in the following table:

50bp
75bp
100bp
150bp
300bp

300x
300_50
300_75
300_100
300_150
300_300

700x
700_50
700_75
700_100
700_150
700_300

1000x
1000_50
1000_75
1000_100
1000_150
1000_300

3000x
3000_50
3000_75
3000_100
3000_150
3000_300

5000x
5000_50
5000_75
5000_100
5000_150
5000_300

This expanded set builds upon our previous dataset release, which varied each parameter independently. Here, we provide a grid design, allowing researchers to analyze both individual and combined effects of coverage and read length on somatic variant calling tools. The files included are aligned files with their index (BAM and BAI) along with the VCF files containing the variants for each sample.

For more information about the synthetic data generation framework, please visit the synth4bench GitHub repositoryOur work has been submitted to the bioRxiv preprint repository. If you use synth4bench, or any of our data/scripts/code, please cite doi:10.1101/2024.03.07.582313v2.