Dupre

Dupre estimates the duplicate rate of sequencing libraries at specified sequencing depths using occupancy vectors derived from small subsamples to provide a quality metric for high-throughput sequencing (HTS).

Key Features:

• Duplicate-rate estimation: Provides an estimate of the true duplicate rate from small, low-depth subsamples, including multiplexed libraries.
• Occupancy-vector input: Uses known occupancy vectors from subsamples as the primary input data.
• Up-sampling of occupancy distribution: Performs up-sampling of the occupancy distribution of reads' copy numbers to predict duplicate rates at higher sequencing depths.
• Mathematical model: Implements an explicit, elementary mathematical framework that inverts the sub-sampling process for interpretability.
• Optimization methods: Employs quadratic and linear optimization for computational estimation.
• HTS applicability: Targets sequencing libraries and specified sequencing depths within high-throughput sequencing workflows.
• Benchmarking: Has been evaluated on artificial and real datasets against previous approaches.
• Generalisability: Underlying principles can be extended for general diversity estimation tasks beyond duplicate-rate assessment.

Scientific Applications:

• Sequencing quality control: Estimating library duplicate rates as a quality metric for HTS projects.
• Experimental planning: Predicting duplicate rates at target sequencing depths to inform sequencing strategy and resource allocation.
• Library preparation diagnostics: Identifying issues such as low input DNA or excessive PCR cycles through observed duplicate-rate patterns.
• Diversity estimation: Extending the methodology to broader diversity estimation problems beyond duplicate-rate measurement.

Methodology:

Dupre uses an elementary mathematical framework that explicitly inverts the sub-sampling process by up-sampling the occupancy distribution of reads' copy numbers and performs computational estimation via quadratic and linear optimization.