A Bayesian hierarchical model for inference across related reverse phase protein arrays experiments

Riten Mitra, Peter Müller, Yuan Ji, Yitan Zhu, Gordon Mills, Yiling Lu

Research output: Contribution to journalArticle

Abstract

We consider inference for functional proteomics experiments that record protein activation over time following perturbation under different dose levels of several drugs. The main inference goal is the dependence structure of the selected proteins. A critical challenge is the lack of sufficient data under any one drug and dose level to allow meaningful inference on dependence structure. We propose a hierarchical model to implement the desired inference. The key element of the model is a shared dependence structure on (latent) binary indicators of protein activation.

Original languageEnglish (US)
Pages (from-to)2483-2492
Number of pages10
JournalJournal of Applied Statistics
Volume41
Issue number11
DOIs
StatePublished - Jan 1 2014
Externally publishedYes

Fingerprint

Bayesian Hierarchical Model
Reverse
Dependence Structure
Protein
Experiment
Activation
Dose
Drugs
Proteomics
Hierarchical Model
Binary
Sufficient
Perturbation
Inference
Bayesian hierarchical model
Dependence structure

Keywords

  • Bayesian
  • graphical
  • hierarchical model
  • protein networks
  • timecourse

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

A Bayesian hierarchical model for inference across related reverse phase protein arrays experiments. / Mitra, Riten; Müller, Peter; Ji, Yuan; Zhu, Yitan; Mills, Gordon; Lu, Yiling.

In: Journal of Applied Statistics, Vol. 41, No. 11, 01.01.2014, p. 2483-2492.

Research output: Contribution to journalArticle

Mitra, Riten ; Müller, Peter ; Ji, Yuan ; Zhu, Yitan ; Mills, Gordon ; Lu, Yiling. / A Bayesian hierarchical model for inference across related reverse phase protein arrays experiments. In: Journal of Applied Statistics. 2014 ; Vol. 41, No. 11. pp. 2483-2492.
@article{61ff5c8019294ce5becc09e51a3ed8aa,
title = "A Bayesian hierarchical model for inference across related reverse phase protein arrays experiments",
abstract = "We consider inference for functional proteomics experiments that record protein activation over time following perturbation under different dose levels of several drugs. The main inference goal is the dependence structure of the selected proteins. A critical challenge is the lack of sufficient data under any one drug and dose level to allow meaningful inference on dependence structure. We propose a hierarchical model to implement the desired inference. The key element of the model is a shared dependence structure on (latent) binary indicators of protein activation.",
keywords = "Bayesian, graphical, hierarchical model, protein networks, timecourse",
author = "Riten Mitra and Peter M{\"u}ller and Yuan Ji and Yitan Zhu and Gordon Mills and Yiling Lu",
year = "2014",
month = "1",
day = "1",
doi = "10.1080/02664763.2014.920776",
language = "English (US)",
volume = "41",
pages = "2483--2492",
journal = "Journal of Applied Statistics",
issn = "0266-4763",
publisher = "Routledge",
number = "11",

}

TY - JOUR

T1 - A Bayesian hierarchical model for inference across related reverse phase protein arrays experiments

AU - Mitra, Riten

AU - Müller, Peter

AU - Ji, Yuan

AU - Zhu, Yitan

AU - Mills, Gordon

AU - Lu, Yiling

PY - 2014/1/1

Y1 - 2014/1/1

N2 - We consider inference for functional proteomics experiments that record protein activation over time following perturbation under different dose levels of several drugs. The main inference goal is the dependence structure of the selected proteins. A critical challenge is the lack of sufficient data under any one drug and dose level to allow meaningful inference on dependence structure. We propose a hierarchical model to implement the desired inference. The key element of the model is a shared dependence structure on (latent) binary indicators of protein activation.

AB - We consider inference for functional proteomics experiments that record protein activation over time following perturbation under different dose levels of several drugs. The main inference goal is the dependence structure of the selected proteins. A critical challenge is the lack of sufficient data under any one drug and dose level to allow meaningful inference on dependence structure. We propose a hierarchical model to implement the desired inference. The key element of the model is a shared dependence structure on (latent) binary indicators of protein activation.

KW - Bayesian

KW - graphical

KW - hierarchical model

KW - protein networks

KW - timecourse

UR - http://www.scopus.com/inward/record.url?scp=84926080013&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84926080013&partnerID=8YFLogxK

U2 - 10.1080/02664763.2014.920776

DO - 10.1080/02664763.2014.920776

M3 - Article

AN - SCOPUS:84926080013

VL - 41

SP - 2483

EP - 2492

JO - Journal of Applied Statistics

JF - Journal of Applied Statistics

SN - 0266-4763

IS - 11

ER -