CellSort

CellSort optimizes fluorescence-activated cell sorting (FACS) gating and predicts enrichment rounds to improve high-throughput protein engineering and directed evolution screening.

Key Features:

• Support Vector Machine (SVM) Algorithm: Uses a support vector machine trained on positive and negative control populations to identify optimal sorting gates.
• Multi-Dimensional Analysis: Operates in more than two dimensions to distinguish cell populations beyond conventional two-dimensional FACS analyses.
• Bayesian Enrichment Prediction: Applies a Bayesian framework to predict the number of sorting rounds required to enrich a population from a given library size.
• Gate Biasing Strategy: Enables biasing of sorting gates based on predictive outputs to reduce the number of enrichment rounds and address false positive/negative rates.

Scientific Applications:

• Protein Engineering: Optimizes FACS-based selection steps in protein engineering workflows.
• Directed Evolution: Reduces enrichment cycles in directed evolution campaigns that rely on FACS screening.
• High-Throughput Screening: Improves gate definition and reduces false positives and negatives in high-throughput FACS screens.
• Multi-Parameter Cell Population Analysis: Enhances discrimination of complex cell populations using multi-dimensional FACS data.

Methodology:

Train an SVM on positive and negative control populations in multi-dimensional feature space to determine optimal sorting gates, and use a Bayesian framework to forecast enrichment rounds and guide gate biasing.