Inferring cell-scale signalling networks via compressive sensing.

Area of Science:

• Systems Biology
• Computational Biology
• Network Inference

Background:

• Inferring biological signalling networks is crucial for understanding cellular processes.
• Previous methods often infer local networks independently, potentially limiting accuracy due to system indivisibility.
• Holistic, cell-scale network inference faces challenges in scalability, measurement, and overfitting.

Purpose of the Study:

• To develop a feasible holistic approach for cell-scale signalling network inference.
• To address the limitations of scalability, measurement, and overfitting in holistic network inference.
• To improve the accuracy of signalling network inference by considering the entire system.

Main Methods:

• Proposed CCELL, a method for cell-scale signalling network inference.
• Utilized Bayesian compressive sensing on time-series data generated by immunoprecipitation.
• Leveraged observations of sparse concentration variations and cross-reactivity for efficient measurement.

Main Results:

• Demonstrated the effectiveness of CCELL using benchmark networks with varying numbers of time-variant species.
• Achieved high accuracy in network inference from a reduced number of measurements.
• Overcame scalability and measurement challenges associated with holistic network inference.

Conclusions:

• CCELL offers a robust and accurate method for inferring cell-scale signalling networks holistically.
• The approach effectively handles sparse variations and enables accurate inference from limited data.
• This work advances systems biology by providing a more accurate and efficient tool for network reconstruction.