Predictive modeling of genome-wide mRNA expression: from modules to molecules

Area of Science:

• Computational biology
• Systems biology
• Genomics

Background:

• Numerous algorithms exist for predicting messenger RNA (mRNA) expression and modeling gene regulatory networks.
• These algorithms vary in their approach, including reliance on coregulated gene modules versus all-gene models, representation of regulatory activities, and explicit modeling of cis-regulatory logic involving transcription factors.

Purpose of the Study:

• To explore integrative computational analysis strategies for gene regulation by combining different types of functional genomics data.
• To present a promising avenue for accurate and comprehensive gene regulation modeling using biophysical principles and large-scale data.

Main Methods:

• Utilizing large-scale functional genomics data.
• Employing biophysical modeling of molecular interactions (proteins, DNA, RNA).
• Estimating regulatory network connectivity and activity parameters through data integration.

Main Results:

• Functional genomics data from various sources reflect common molecular processes, enabling integrative analysis.
• Combining biophysical modeling with functional genomics data is a promising strategy for accurate gene regulation models.
• Advancements in modeling complex cis-regulatory logic may reduce reliance on cross-species conservation approaches.

Conclusions:

• Integrative computational analysis of diverse functional genomics data is key to advancing gene regulation models.
• Biophysical modeling coupled with large-scale data provides a robust framework for understanding gene regulatory networks.
• Future models may increasingly capture complex regulatory logic, potentially diminishing the need for cross-species conservation data.