When researchers use combinatorial genetic engineering, they often rely on next-generation sequencing (NGS) to track and measure barcoded or mutated genes. Choosing how deep to sequence, meaning how many times each piece of DNA is read, is a critical step because it affects accuracy, cost, and data quality.

A team led by researchers from Sorbonne Université set out to tackle this challenge. While established guidelines exist for experiments such as RNA sequencing or whole genome sequencing, no standard rules were available for barcoded DNA libraries, which are often used in high-throughput genetic experiments.

By analyzing real NGS datasets and building a mathematical model that considered how DNA is amplified during library preparation, the researchers discovered something surprising. Increasing sequencing depth does not always improve accuracy. In fact, beyond a certain point, deeper sequencing actually adds more noise to the data instead of clarifying results.

The team proposes a simple rule of thumb. The optimal sequencing depth should be about ten times the number of barcoded DNA molecules present before amplification. This practical guideline could help scientists design more efficient experiments, saving time and resources while still ensuring reliable results.