Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Gene Network Reconstruction using Global-Local Shrinkage Priors.

Gwenaël G R Leday1, Mathisca C M de Gunst2, Gino B Kpogbezan3

  • 1MRC Biostatistics Unit, Cambridge Institute of Public Health, Forvie Site, Robinson Way, Cambridge Biomedical Campus, Cambridge CB2 0SR, United Kingdom.

The Annals of Applied Statistics
|April 15, 2017
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Statescope: an integrative deconvolution framework for discovering cell states in tumors.

Nature communications·2026
Same author

Prognostic value of peri-operative circulating tumour DNA levels estimated by cell-free DNA methylation in patients with resectable colorectal liver metastases.

EBioMedicine·2026
Same author

Signpost Testing to Navigate the Parameter Space of the Gaussian Graphical Model With High-Dimensional Data.

Biometrical journal. Biometrische Zeitschrift·2026
Same author

Informative Co-Data Learning for High-Dimensional Horseshoe Regression.

Biometrical journal. Biometrische Zeitschrift·2025
Same author

Computationally efficient multi-sample flow cytometry data analysis using Gaussian mixture models.

BMC bioinformatics·2025
Same author

False Discovery Estimation in Record Linkage.

Statistics in medicine·2025

This study introduces ShrinkNet, a novel method for reconstructing gene networks from molecular data. ShrinkNet improves gene interaction inference by combining local and global regularization, outperforming existing methods.

Area of Science:

  • Computational Biology
  • Bioinformatics
  • Systems Biology

Background:

  • Reconstructing gene networks from high-throughput molecular data is complex due to a large number of parameters compared to sample size.
  • Traditional methods often use local regularization, which can lead to difficulties in estimating parameters and large statistical uncertainties.

Purpose of the Study:

  • To propose a novel inference strategy that combines local regularization with global shrinkage for improved gene network reconstruction.
  • To enhance the estimation of regularization parameters by borrowing strength across genes.

Main Methods:

  • A simple Bayesian model with non-sparse, conjugate priors was employed.
  • Fast variational approximations to posteriors were utilized.
  • Empirical Bayes estimation of hyper-parameters and a novel rank-based posterior thresholding approach were developed.
Keywords:
Bayesian inferenceEmpirical BayesShrinkageUndirected gene networkVariational approximation

Related Experiment Videos

Main Results:

  • The proposed ShrinkNet method demonstrated superior performance compared to popular sparse methods in simulations.
  • ShrinkNet yielded more stable gene network edges and improved reproducibility.
  • The method was applied to Glioblastoma data to study gene interactions related to patient survival.

Conclusions:

  • ShrinkNet offers a robust and reproducible approach for gene network inference from high-throughput molecular data.
  • The combination of local and global shrinkage effectively addresses challenges in parameter estimation.
  • The application to Glioblastoma highlights the method's utility in biological network analysis.