icbn

icbn implements parameter estimation and model selection algorithms for Isotonic Conjunctive Bayesian Network (I-CBN) models to estimate order constraints among genetic events and map genotype-phenotype relationships.

Key Features:

• Isotonic Regression Integration: Applies isotonic regression on partially ordered sets (the lattice of genotypes) to enforce a non-decreasing genotype-phenotype map.
• Conjunctive Bayesian Networks (CBNs): Uses conjunctive Bayesian networks to represent partial orders of genetic events and model dependencies among accumulating mutations.
• Parameter Estimation and Model Selection: Implements algorithms for parameter estimation and model selection within the I-CBN framework.

Scientific Applications:

• Evolutionary Dynamics Analysis: Models accumulation of mutations to study pathogen escape mechanisms from immune responses and medical interventions.
• Genotype-Phenotype Mapping: Estimates genotype-phenotype relationships to relate genetic variation to observable traits or fitness levels.
• Drug Resistance Studies: Applied to HIV drug resistance data to identify non-linear effects of resistance mutations and their dependence on genetic background.

Methodology:

Integrates isotonic regression with conjunctive Bayesian networks to capture order constraints and dependencies among genetic events and validates genotype-phenotype maps using simulated data.