BIGL

BIGL evaluates drug combination synergy by comparing observed combination effects to expected outcomes under specified null models to support early-stage drug discovery.

Key Features:

• Synergy Evaluation: Screens drug combinations for synergy by comparing observed effects to predictions from null models including Loewe, Loewe2, HSA, and Bliss.
• Extended Statistical Tests: Extends meanR and maxR tests to accommodate non-constant (heteroscedastic) variance in response data.
• Null Model Flexibility: Allows selection among multiple null models to match expert knowledge and experimental assumptions.
• Comprehensive Simulation Study: Evaluates performance via extensive simulations under different additivity and synergy models and monotherapeutic dose-response scenarios (complete, partial, and incomplete responders), including deviations from the constant variance assumption.
• Bootstrap Confidence Intervals: Provides bootstrap confidence intervals for individual off-axis points and for overall synergy strength.

Scientific Applications:

• Early-stage drug discovery: Identification and prioritization of synergistic, additive, or antagonistic drug pairs for follow-up in combination therapy research.
• Method evaluation and robustness assessment: Benchmarking and assessing statistical properties of synergy detection across heteroscedastic responses and varying monotherapeutic dose-response types.

Methodology:

Comparison of observed combination effects to expected outcomes derived from selected null models (Loewe, Loewe2, HSA, Bliss); extension of meanR and maxR tests for non-constant variance; bootstrap confidence intervals for off-axis points and overall synergy strength; performance assessment via simulation studies under specified additivity/synergy models and monotherapeutic response scenarios (complete, partial, incomplete) including variance deviations.