Related Experiment Video
Updated: Apr 19, 2026
03:08
Using Human Differentially Expressed Gene Lists to Perform Downstream Pathway Enrichment Analysis and Target Prioritization
Published on: October 3, 2025
1.1K
SMARTS: reconstructing disease response networks from multiple individuals using time series gene expression data.
1Lane Center for Computational Biology and Machine Learning Department, Carnegie Mellon University, Pittsburgh, PA, USA.
Bioinformatics (Oxford, England)
|December 7, 2014
Summary
Scalable Models for the Analysis of Regulation from Time Series (SMARTS) reconstructs condition-specific regulatory networks from multiple individuals. This method improves groupings and identifies key transcription factors differentiating responses in human and mouse models.
Area of Science:
- Computational Biology
- Systems Biology
- Bioinformatics
Background:
- Current methods for dynamic regulatory network reconstruction primarily use model organisms or cell lines, which do not account for human variability.
- Human time-series data exhibit individual differences in background expression, start/end times, and response rates, necessitating new integration methods.
- Reconstructing disease-specific regulatory networks requires methods that can handle multi-individual time-series data.
Purpose of the Study:
- To develop a novel computational method for reconstructing condition-specific dynamic regulatory networks from multi-individual time-series data.
- To address the challenges of human biological variability, including differing expression profiles, start times, and response rates.
- To enable unsupervised grouping of individuals and identification of key regulatory factors differentiating these groups.
Main Methods:
- Developed Scalable Models for the Analysis of Regulation from Time Series (SMARTS), a method integrating static and time-series data from multiple individuals.
- Employed probabilistic graphical models to iteratively reconstruct regulatory networks and assign individuals, accounting for individual variations in timing and response rates.
- Applied an unsupervised approach to identify condition-specific response networks.
Main Results:
- SMARTS successfully integrated multi-individual data to reconstruct condition-specific regulatory networks.
- The method improved baseline groupings and identified key transcription factors (TFs) that differentiate responses in human influenza and mouse brain development datasets.
- Identified both known and novel TFs involved in immune response and development, providing new biological hypotheses.
Conclusions:
- SMARTS is an effective method for reconstructing dynamic regulatory networks from diverse, multi-individual human time-series data.
- The approach facilitates the discovery of patient subgroups and their specific regulatory mechanisms.
- SMARTS offers a powerful tool for understanding biological responses in complex systems and for hypothesis generation in disease and development.
