SeqControl

SeqControl predicts sequencing quality and required sequencing depth from whole-genome sequencing metrics to forecast coverage and basewise quality for high-throughput sequencing quality control.

Key Features:

• Coverage and quality metrics: Uses a set of 15 metrics describing overall coverage, coverage distribution, basewise coverage, and basewise quality.
• Predictive modeling: Employs multivariate models derived from whole-genome sequences of 27 prostate cancers and 26 normal references.
• Sequencing depth forecasting: Predicts the amount of sequencing required to achieve a specified coverage depth with reported AUC = 0.993.
• Clinical sample classification: Classifies clinically relevant samples, including formalin-fixed paraffin-embedded (FFPE) specimens, from as little as one-eighth of a sequencing lane with reported AUC = 0.967.
• Implementation: Implemented using Perl and R scripts.

Scientific Applications:

• Quality control in high-throughput sequencing: Forecasts coverage and basewise quality to inform sequencing QC decisions.
• Sequencing depth optimization: Guides allocation of sequencing resources by predicting required sequencing to reach target coverage depths.
• Clinical sample assessment: Enables classification of clinical and FFPE samples from limited sequencing data.

Methodology:

Computes 15 coverage and basewise quality metrics and trains multivariate predictive models on whole-genome sequences from 27 prostate cancers and 26 normal references; implemented in Perl and R scripts.