BAYESIAN FEATURE SELECTION WITH DATA INTEGRATION

Biomarker discovery aims to select biological markers that differentiate between different groups, and is typically framed as a small-sample high-dimensional feature selection problem. Successfully combining information from multiple datasets studying similar diseases under comparable conditions can increase the accuracy of biomarker discovery. Recent work proposes a Bayesian feature selection framework that finds the sample-conditioned probability of a feature having distributional differences across classes. Here we extend this framework to consider multiple datasets, solve it for Gaussian features, and extend two previously proposed selection algorithms. The proposed algorithms perform well on synthetic simulations. We use them to integrate four breast cancer datasets and perform enrichment analysis.

• Publication year

  2018

• Venue

  IEEE Global Conference on Signal and Information Processing

• Publication date

  2018-11-01

• Fields of study

  Biology, Medicine, Computer Science

• Identifiers
  DOI 10.1109/GlobalSIP.2018.8646612
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• High Dimensional Data Enrichment: Interpretable, Fast, and Data-Efficient

  2018cited by this paper
• Robust feature selection for block covariance Bayesian models

  2017cited by this paper
• Heuristic algorithms for feature selection under Bayesian models with block-diagonal covariance structure

  2017cited by this paper
• Discordancy Partitioning for Validating Potentially Inconsistent Pharmacogenomic Studies

  2017cited by this paper
• Optimal Bayesian feature filtering for single-nucleotide polymorphism data

  2017cited by this paper
• Multiclass Bayesian feature selection

  2017cited by this paper
• Data Shared Lasso: A novel tool to discover uplift

  2016cited by this paper
• Optimal Bayesian feature selection with missing data

  2016cited by this paper
• PANTHER version 11: expanded annotation data from Gene Ontology and Reactome pathways, and data analysis tool enhancements

  2016influential reference
• Microarray Meta-Analysis and Cross-Platform Normalization: Integrative Genomics for Robust Biomarker Discovery

  2015influential reference
• Optimal bayesian feature filtering

  2015cited by this paper
• A genomic predictor of response and survival following taxane-anthracycline chemotherapy for invasive breast cancer.

  2011cited by this paper
• Performance of feature-selection methods in the classification of high-dimension data

  2009cited by this paper
• PANTHER pathway: an ontology-based pathway database coupled with data analysis tools.

  2009cited by this paper
• A gene expression signature that defines breast cancer metastases

  2008cited by this paper
• Supervised feature selection via dependence estimation

  2007influential reference
• Strong Time Dependence of the 76-Gene Prognostic Signature for Node-Negative Breast Cancer Patients in the TRANSBIG Multicenter Independent Validation Series

  2007cited by this paper
• Bayesian variable selection and data integration for biological regulatory networks

  2006cited by this paper
• Gene-expression profiles to predict distant metastasis of lymph-node-negative primary breast cancer.

  2005cited by this paper
• Gene expression profiling spares early breast cancer patients from adjuvant therapy: derived and validated in two population-based cohorts

  2005cited by this paper
• Biomarker discovery and validation: technologies and integrative approaches.

  2004cited by this paper
• Gene Expression Omnibus: NCBI gene expression and hybridization array data repository

  2002cited by this paper
• Host microenvironment in breast cancer development: Inflammatory cells, cytokines and chemokines in breast cancer progression: reciprocal tumor–microenvironment interactions

  2002cited by this paper
• Nonparametric entropy estimation. An overview

  1997cited by this paper

• Binary Classification for Failure Risk Assessment.

  2021cites this paper