Missing data imputation using a truncated Gaussian infinite factor model with application to metabolomics data

Metabolomics is the study of small molecules in biological samples. Metabolomics data are typically high-dimensional and contain highly correlated variables and frequent missing values. Both missing at random (MAR) data, due to acquisition or processing errors, and missing not at random (MNAR) data, caused by values falling below detection thresholds, are common. Thus, imputation is a critical component of downstream analysis. Existing imputation methods generally assume one type of data missingness mechanism, or impute values outside the data's physical constraints. A novel truncated Gaussian infinite factor analysis (TGIFA) model is proposed to perform statistically principled and physically realistic imputation in metabolomics data. By incorporating truncated Gaussian assumptions, TGIFA respects the data's physical constraints, while leveraging an infinite latent factor framework to capture high-dimensional dependencies without pre-specifying the number of latent factors. Our Bayesian inference approach enables uncertainty quantification in both the values of the imputed data, and the missing data mechanism. A computationally efficient exchange algorithm enables scalable posterior inference via Markov Chain Monte Carlo. We validate TGIFA through a comprehensive simulation study and demonstrate its utility in a motivating urinary metabolomics dataset, where it yields useful imputations, with associated uncertainty quantification. Open-source R code, available at https://github.com/kfinucane/TGIFA, accompanies TGIFA.

• Publication year

  2024

• Venue

  Unknown venue

• Publication date

  2024-10-14

• Fields of study

  Mathematics, Chemistry

• Identifiers
  arXiv 2410.10633
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Exploration of differential responses to FODMAPs and gluten in people with irritable bowel syndrome- a double-blind randomized cross-over challenge study

  2024cited by this paper
• Clustered Mallows Model

  2024cited by this paper
• MetaboAnalyst 6.0: towards a unified platform for metabolomics data processing, analysis and interpretation

  2024cited by this paper
• Efficiency Assessment and Determinants of Performance: A Study of Jordan’s Banks Using DEA and Tobit Regression

  2024cited by this paper
• Metabolomic based approach to identify biomarkers of broccoli intake

  2023cited by this paper
• Pretreating and normalizing metabolomics data for statistical analysis

  2023cited by this paper
• Generalized cumulative shrinkage process priors with applications to sparse Bayesian factor analysis

  2023cited by this paper
• Untargeted metabolomics by liquid chromatography-mass spectrometry for food authentication: A review.

  2022cited by this paper
• Imputation of Missing Values for Multi-Biospecimen Metabolomics Studies: Bias and Effects on Statistical Validity

  2022influential reference
• A Latent Shrinkage Position Model for Binary and Count Network Data

  2022cited by this paper
• Metabolomics in infectious diseases and drug discovery

  2021cited by this paper
• Generalized infinite factorization models.

  2021cited by this paper
• Multi-omics integration in biomedical research - A metabolomics-centric review.

  2021cited by this paper
• Applications of Metabolomics to Precision Nutrition

  2021cited by this paper
• Comparison of imputation and imputation-free methods for statistical analysis of mass spectrometry data with missing data

  2021cited by this paper
• Determination of safe limit for arsenic contaminated irrigation water using solubility free ion activity model (FIAM) and Tobit Regression Model.

  2020cited by this paper
• Robust sparse Bayesian infinite factor models

  2020cited by this paper
• Inferring food intake from multiple biomarkers using a latent variable model

  2020cited by this paper
• BayesMetab: treatment of missing values in metabolomic studies using a Bayesian modeling approach

  2019cited by this paper
• Bayesian cumulative shrinkage for infinite factorizations.

  2019cited by this paper
• Data analysis of MS-based clinical lipidomics studies with crossover design: A tutorial mini-review of statistical methods.

  2019cited by this paper
• Classification and identification of molecules through factor analysis method based on terahertz spectroscopy.

  2018cited by this paper
• Flexible Imputation of Missing Data

  2018cited by this paper
• Sparse Bayesian Factor Analysis When the Number of Factors Is Unknown

  2018cited by this paper
• Missing Value Imputation Approach for Mass Spectrometry-based Metabolomics Data

  2018cited by this paper
• Infinite Mixtures of Infinite Factor Analysers

  2017cited by this paper
• Navigating freely-available software tools for metabolomics analysis

  2017cited by this paper
• GSimp: A Gibbs sampler based left-censored missing value imputation approach for metabolomics studies

  2017cited by this paper
• Distribution based nearest neighbor imputation for truncated high dimensional data with applications to pre-clinical and clinical metabolomics studies

  2017cited by this paper
• Power Analysis and Sample Size Determination in Metabolic Phenotyping.

  2016cited by this paper
• Clustering with the multivariate normal inverse Gaussian distribution

  2016cited by this paper
• A note on the multiplicative gamma process

  2016cited by this paper
• Dimension reduction techniques for the integrative analysis of multi-omics data

  2016cited by this paper
• missForest: Nonparametric missing value imputation using random forest

  2015cited by this paper
• Influence of Missing Values Substitutes on Multivariate Analysis of Metabolomics Data

  2014cited by this paper
• Metabolomic fingerprinting: challenges and opportunities.

  2013cited by this paper
• Bayesian Structure Learning in Sparse Gaussian Graphical Models

  2012cited by this paper
• Multivariate Analysis in Metabolomics.

  2012cited by this paper
• Sparse Bayesian infinite factor models.

  2011influential reference
• MissForest - non-parametric missing value imputation for mixed-type data

  2011cited by this paper
• Multivariate methods in metabolomics – from pre-processing to dimension reduction and statistical analysis

  2011cited by this paper
• Latent Variable Models and Factor Analysis: A Unified Approach

  2011cited by this paper
• Nonparametric Bayesian Sparse Factor Models with application to Gene Expression modelling

  2010cited by this paper
• Simulation of truncated normal variables

  2009cited by this paper
• Parsimonious Gaussian mixture models

  2008cited by this paper
• Pattern Recognition and Machine Learning

  2006cited by this paper
• Large-scale human metabolomics studies: a strategy for data (pre-) processing and validation.

  2006cited by this paper
• MCMC for Doubly-intractable Distributions

  2006cited by this paper
• On ranking and selection from independent truncated normal distributions

  2005cited by this paper
• Some results on the multivariate truncated normal distribution

  2005cited by this paper
• Efficient Gibbs Sampling of Truncated Multivariate Normal with Application to Constrained Linear Regression

  2004cited by this paper
• A Bayesian missing value estimation method for gene expression profile data

  2003cited by this paper
• Gibbs sampling approach for generation of truncated multivariate Gaussian random variables

  1999cited by this paper
• An Introduction to Variational Methods for Graphical Models

  1999cited by this paper
• Normal Inverse Gaussian Distributions and Stochastic Volatility Modelling

  1997cited by this paper
• Hybrid Monte Carlo

  1988cited by this paper

• No citing papers are available for this paper.