MIIC online: a web server to reconstruct causal or non-causal networks from non-perturbative data

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Network reconstruction from observational data is crucial for understanding complex biological systems.
• Existing methods often require a priori assumptions about network structure (causal vs. non-causal).

Purpose of the Study:

• To present a novel web server implementing the MIIC algorithm for robust network reconstruction.
• To enable the identification of causal, non-causal, or mixed networks without prior assumptions.

Main Methods:

• The MIIC (Mutual Information and Information Content) algorithm combines constraint-based and information-theoretic frameworks.
• It iteratively removes dispensable edges by subtracting information contributions from indirect paths.
• Edges are filtered by confidence or oriented using causality signatures in observational data.

Main Results:

• The MIIC algorithm reconstructs networks from non-perturbative data, accommodating potential latent variables.
• It demonstrates superior or comparable performance to state-of-the-art methods for both causal and non-causal network reconstruction.
• Applicable to diverse biological data, including single-cell gene expression and protein sequence evolution.

Conclusions:

• The MIIC online server provides a versatile tool for biological network inference.
• It advances the field by offering a unified approach to causal and non-causal network reconstruction.
• The method's flexibility and performance make it valuable for analyzing complex biological data.