Stochastic Boolean networks: an efficient approach to modeling gene regulatory networks

Area of Science:

• Computational biology
• Systems biology
• Bioinformatics

Background:

• Probabilistic Boolean networks (PBNs) are valuable for modeling gene regulatory networks (GRNs) and understanding their dynamics.
• PBNs incorporate molecular and genetic noise, crucial for disease mechanism insights and therapeutic development.
• However, PBNs face computational complexity in calculating state transition matrices and steady-state distributions.

Purpose of the Study:

• To introduce a novel, efficient implementation of PBNs using stochastic logic and computation, termed Stochastic Boolean Networks (SBNs).
• To demonstrate the capability of SBNs in accurately simulating PBNs, including responses to gene perturbations.
• To enable efficient analysis of PBN steady-state distributions.

Main Methods:

• Developed a stochastic implementation of PBNs (SBNs) based on stochastic logic and computation.
• Implemented a time-frame expanded SBN for efficient steady-state distribution analysis.
• Validated SBNs through simulations on p53, random, and T cell immune response networks.

Main Results:

• SBNs compute the state transition matrix with a complexity of O(nL2n), which is more efficient than traditional PBNs, especially for large networks.
• SBNs provide accurate simulations of PBNs, both with and without gene perturbations.
• Efficient analysis of steady-state distributions is achieved using a time-frame expanded SBN.

Conclusions:

• Stochastic Boolean networks (SBNs) provide an efficient and accurate method for modeling gene regulatory networks (GRNs).
• SBNs successfully replicate known biological behaviors, such as p53-Mdm2 network oscillations and T cell immune response attractors.
• The SBN approach offers valuable insights into GRN dynamics under perturbation, aiding in understanding disease mechanisms.