BNPower

BNPower performs power calculation for data-driven network analysis of whole-brain connectome data using simulation analysis that accounts for the complexity of data-driven subnetwork identification, multivariate edge variables, and multiple testing correction.

Key Features:

• Simulation-based power calculation: BNPower utilizes simulation analysis to address the challenge of power calculation for data-driven network analysis where nodes, edges, and model parameters cannot be pre-specified before analysis.
• Subnetwork extraction: BNPower implements efficient computational strategies for extracting subnetworks from connectome data represented as matrices.
• Multiple testing correction: BNPower incorporates permutation tests for controlling family-wise error rate (FWER) while maintaining low computational costs.
• Multivariate edge variable handling: BNPower handles analyses involving multivariate edge variables and multiple subnetworks.

Scientific Applications:

• Connectome network analysis: BNPower enables power calculation for examining systematic changes in connections among brain areas caused by clinical and experimental conditions.
• Data-driven subnetwork identification: BNPower supports power analysis for approaches that automatically organize individual predictor-related edges into subnetworks rather than relying on pre-specified subnetworks.
• Network-level inference: BNPower facilitates power calculation for network-level inference in whole-brain connectome studies where connectome data are treated as outcomes and subject conditions serve as predictors.

Methodology:

BNPower uses simulation analysis incorporating subnetwork extraction and permutation tests for family-wise error rate control.