Fast zero-inflated negative binomial mixed modeling approach for analyzing longitudinal metagenomics data

2020 ◽  
Vol 36 (8) ◽  
pp. 2345-2351 ◽  
Author(s):  
Xinyan Zhang ◽  
Nengjun Yi
Keyword(s):  
Count Data ◽  
Mixed Models ◽  
Negative Binomial ◽  
Linear Mixed Models ◽  
Human Microbiome ◽  
Real Data ◽  
R Package ◽  
Supplementary Information ◽  
Sequencing Data ◽  
Metagenomics Data

Abstract Motivation Longitudinal metagenomics data, including both 16S rRNA and whole-metagenome shotgun sequencing data, enhanced our abilities to understand the dynamic associations between the human microbiome and various diseases. However, analytic tools have not been fully developed to simultaneously address the main challenges of longitudinal metagenomics data, i.e. high-dimensionality, dependence among samples and zero-inflation of observed counts. Results We propose a fast zero-inflated negative binomial mixed modeling (FZINBMM) approach to analyze high-dimensional longitudinal metagenomic count data. The FZINBMM approach is based on zero-inflated negative binomial mixed models (ZINBMMs) for modeling longitudinal metagenomic count data and a fast EM-IWLS algorithm for fitting ZINBMMs. FZINBMM takes advantage of a commonly used procedure for fitting linear mixed models, which allows us to include various types of fixed and random effects and within-subject correlation structures and quickly analyze many taxa. We found that FZINBMM remarkably outperformed in computational efficiency and was statistically comparable with two R packages, GLMMadaptive and glmmTMB, that use numerical integration to fit ZINBMMs. Extensive simulations and real data applications showed that FZINBMM outperformed other previous methods, including linear mixed models, negative binomial mixed models and zero-inflated Gaussian mixed models. Availability and implementation FZINBMM has been implemented in the R package NBZIMM, available in the public GitHub repository http://github.com//nyiuab//NBZIMM. Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals scBatch: batch-effect correction of RNA-seq data through sample distance matrix adjustment

2020 ◽  
Vol 36 (10) ◽  
pp. 3115-3123 ◽  
Author(s):  
Teng Fei ◽  
Tianwei Yu
Keyword(s):  
Single Cell ◽  
Differential Expression Analysis ◽  
Distance Matrix ◽  
Real Data ◽  
R Package ◽  
Batch Effect ◽  
Supplementary Information ◽  
Rna Seq ◽  
Sequencing Data ◽  
Gene Differential Expression

Abstract Motivation Batch effect is a frequent challenge in deep sequencing data analysis that can lead to misleading conclusions. Existing methods do not correct batch effects satisfactorily, especially with single-cell RNA sequencing (RNA-seq) data. Results We present scBatch, a numerical algorithm for batch-effect correction on bulk and single-cell RNA-seq data with emphasis on improving both clustering and gene differential expression analysis. scBatch is not restricted by assumptions on the mechanism of batch-effect generation. As shown in simulations and real data analyses, scBatch outperforms benchmark batch-effect correction methods. Availability and implementation The R package is available at github.com/tengfei-emory/scBatch. The code to generate results and figures in this article is available at github.com/tengfei-emory/scBatch-paper-scripts. Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals SPARSim single cell: a count data simulator for scRNA-seq data

2019 ◽  
Author(s):  
Giacomo Baruzzo ◽  
Ilaria Patuzzi ◽  
Barbara Di Camillo
Keyword(s):  
Single Cell ◽  
Count Data ◽  
Simulated Data ◽  
Real Data ◽  
R Package ◽  
Supplementary Information ◽  
Rna Seq ◽  
Distribution Of Zeros ◽  
New Methods ◽  
Research Fields

Abstract Motivation Single cell RNA-seq (scRNA-seq) count data show many differences compared with bulk RNA-seq count data, making the application of many RNA-seq pre-processing/analysis methods not straightforward or even inappropriate. For this reason, the development of new methods for handling scRNA-seq count data is currently one of the most active research fields in bioinformatics. To help the development of such new methods, the availability of simulated data could play a pivotal role. However, only few scRNA-seq count data simulators are available, often showing poor or not demonstrated similarity with real data. Results In this article we present SPARSim, a scRNA-seq count data simulator based on a Gamma-Multivariate Hypergeometric model. We demonstrate that SPARSim allows to generate count data that resemble real data in terms of count intensity, variability and sparsity, performing comparably or better than one of the most used scRNA-seq simulator, Splat. In particular, SPARSim simulated count matrices well resemble the distribution of zeros across different expression intensities observed in real count data. Availability and implementation SPARSim R package is freely available at http://sysbiobig.dei.unipd.it/? q=SPARSim and at https://gitlab.com/sysbiobig/sparsim. Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals NBZIMM: negative binomial and zero-inflated mixed models, with application to microbiome/metagenomics data analysis

2020 ◽  
Vol 21 (1) ◽  
Author(s):  
Xinyan Zhang ◽  
Nengjun Yi
Keyword(s):  
Data Analysis ◽  
Mixed Models ◽  
Negative Binomial ◽  
Longitudinal Design ◽  
R Package ◽  
Metagenomic Data ◽  
The Public ◽  
Metagenomics Data ◽  
R Packages ◽  
Over Dispersion

Abstract Background Microbiome/metagenomic data have specific characteristics, including varying total sequence reads, over-dispersion, and zero-inflation, which require tailored analytic tools. Many microbiome/metagenomic studies follow a longitudinal design to collect samples, which further complicates the analysis methods needed. A flexible and efficient R package is needed for analyzing processed multilevel or longitudinal microbiome/metagenomic data. Results NBZIMM is a freely available R package that provides functions for setting up and fitting negative binomial mixed models, zero-inflated negative binomial mixed models, and zero-inflated Gaussian mixed models. It also provides functions to summarize the results from fitted models, both numerically and graphically. The main functions are built on top of the commonly used R packages nlme and MASS, allowing us to incorporate the well-developed analytic procedures into the framework for analyzing over-dispersed and zero-inflated count or proportion data with multilevel structures (e.g., longitudinal studies). The statistical methods and their implementations in NBZIMM particularly address the data characteristics and the complex designs in microbiome/metagenomic studies. The package is freely available from the public GitHub repository https://github.com/nyiuab/NBZIMM. Conclusion The NBZIMM package provides useful tools for complex microbiome/metagenomics data analysis.

scholarly journals Analysis of correlated count data using generalised linear mixed models exemplified by field data on aggressive behaviour of boars

2015 ◽  
Vol 57 (1) ◽  
pp. 1-19
Author(s):  
Norbert Mielenz ◽  
Joachim Spilke ◽  
Eberhard von Borell
Keyword(s):  
Negative Binomial Distribution ◽  
Count Data ◽  
Binomial Distribution ◽  
Mixed Models ◽  
Aggressive Behaviour ◽  
Negative Binomial ◽  
Linear Mixed Models ◽  
Model Parameters ◽  
Generalised Linear Mixed Models ◽  
Animal Effects

Population-averaged and subject-specific models are available to evaluate count data when repeated observations per subject are present. The latter are also known in the literature as generalised linear mixed models (GLMM). In GLMM repeated measures are taken into account explicitly through random animal effects in the linear predictor. In this paper the relevant GLMMs are presented based on conditional Poisson or negative binomial distribution of the response variable for given random animal effects. Equations for the repeatability of count data are derived assuming normal distribution and logarithmic gamma distribution for the random animal effects. Using count data on aggressive behaviour events of pigs (barrows, sows and boars) in mixed-sex housing, we demonstrate the use of the Poisson »log-gamma intercept«, the Poisson »normal intercept« and the »normal intercept« model with negative binomial distribution. Since not all count data can definitely be seen as Poisson or negative-binomially distributed, questions of model selection and model checking are examined. Emanating from the example, we also interpret the least squares means, estimated on the link as well as the response scale. Options provided by the SAS procedure NLMIXED for estimating model parameters and for estimating marginal expected values are presented.

scholarly journals COMPARATIVE STUDY OF CATTLE TICK RESISTANCE USING GENERALIZED LINEAR MIXED MODELS

2019 ◽  
Vol 37 (1) ◽  
pp. 41
Author(s):  
Amanda Marchi MAIORANO ◽  
Thiago Santos MOTA ◽  
Ana Carolina VERDUGO ◽  
Ricardo Antonio da Silva FARIA ◽  
Beatriz Pressi Molina da SILVA ◽  
...  
Keyword(s):  
Negative Binomial Distribution ◽  
Count Data ◽  
Binomial Distribution ◽  
Mixed Models ◽  
Negative Binomial ◽  
Bos Taurus ◽  
Generalized Linear Mixed Models ◽  
Linear Mixed Models ◽  
Tick Count ◽  
Tick Resistance

Comparison of tick resistance in Bos taurus indicus (Nelore) and Bos taurus taurus (Simmental and Caracu) subspecies was investigated utilizing generalized linear mixed models (GLMMs) with Poisson and Negative binomial distributions. Nelore animals (NE) are known to present greater resistance than t. taurus. Difference between tick resistance in Simmental (SI) and Caracu (CA) breeds has never been reported previously. Three artificial tick infestations were conducted to evaluate tick resistance in these breeds. The statistic point of the present study was to show alternative models for the evaluation of tick count data, the GLMMs. Analysis for tick resistance by GLMM with Negative binomial distribution has never been assessed previously. The analyses were performed by the use of the PROC GLIMMIX procedure of the SAS program. The results showed that GLMM with Negative binomial distribution is appropriated to evaluate tick count data with excess of zero observations avoiding overdispersion problems. Finally, considering multiple comparisons with the Bonferroni test, different pattern of tick infestation was observed for the studied breeds, suggesting that NE is the most resistant breed followed by CA.

scholarly journals SPsimSeq: semi-parametric simulation of bulk and single-cell RNA-sequencing data

2020 ◽  
Vol 36 (10) ◽  
pp. 3276-3278 ◽  
Author(s):  
Alemu Takele Assefa ◽  
Jo Vandesompele ◽  
Olivier Thas
Keyword(s):  
Single Cell ◽  
Rna Sequencing ◽  
Real Data ◽  
Simulation Method ◽  
R Package ◽  
Supplementary Information ◽  
Expression Data ◽  
Sequencing Data ◽  
Wide Range ◽  
Single Cell Rna Sequencing

Abstract Summary SPsimSeq is a semi-parametric simulation method to generate bulk and single-cell RNA-sequencing data. It is designed to simulate gene expression data with maximal retention of the characteristics of real data. It is reasonably flexible to accommodate a wide range of experimental scenarios, including different sample sizes, biological signals (differential expression) and confounding batch effects. Availability and implementation The R package and associated documentation is available from https://github.com/CenterForStatistics-UGent/SPsimSeq. Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals Analysis of correlated count data using generalised linear mixed models exemplified by field data on aggressive behaviour of boars

2015 ◽  
Vol 57 (1) ◽  
pp. 1-19
Author(s):  
Norbert Mielenz ◽  
Joachim Spilke ◽  
Eberhard von Borell
Keyword(s):  
Negative Binomial Distribution ◽  
Count Data ◽  
Binomial Distribution ◽  
Mixed Models ◽  
Aggressive Behaviour ◽  
Negative Binomial ◽  
Linear Mixed Models ◽  
Model Parameters ◽  
Generalised Linear Mixed Models ◽  
Animal Effects

Abstract. Population-averaged and subject-specific models are available to evaluate count data when repeated observations per subject are present. The latter are also known in the literature as generalised linear mixed models (GLMM). In GLMM repeated measures are taken into account explicitly through random animal effects in the linear predictor. In this paper the relevant GLMMs are presented based on conditional Poisson or negative binomial distribution of the response variable for given random animal effects. Equations for the repeatability of count data are derived assuming normal distribution and logarithmic gamma distribution for the random animal effects. Using count data on aggressive behaviour events of pigs (barrows, sows and boars) in mixed-sex housing, we demonstrate the use of the Poisson »log-gamma intercept«, the Poisson »normal intercept« and the »normal intercept« model with negative binomial distribution. Since not all count data can definitely be seen as Poisson or negative-binomially distributed, questions of model selection and model checking are examined. Emanating from the example, we also interpret the least squares means, estimated on the link as well as the response scale. Options provided by the SAS procedure NLMIXED for estimating model parameters and for estimating marginal expected values are presented.

scholarly journals Transition models for count data: a flexible alternative to fixed distribution models

2021 ◽  
Author(s):  
Moritz Berger ◽  
Gerhard Tutz
Keyword(s):  
Count Data ◽  
Regression Models ◽  
Negative Binomial ◽  
Real Data ◽  
Distribution Models ◽  
Explanatory Variables ◽  
Excess Zeros ◽  
Proposed Model ◽  
Transition Models ◽  
Fixed Distribution

AbstractA flexible semiparametric class of models is introduced that offers an alternative to classical regression models for count data as the Poisson and Negative Binomial model, as well as to more general models accounting for excess zeros that are also based on fixed distributional assumptions. The model allows that the data itself determine the distribution of the response variable, but, in its basic form, uses a parametric term that specifies the effect of explanatory variables. In addition, an extended version is considered, in which the effects of covariates are specified nonparametrically. The proposed model and traditional models are compared in simulations and by utilizing several real data applications from the area of health and social science.

Detection of differentially methylated CpG sites between tumor samples with uneven tumor purities

2019 ◽  
Vol 36 (7) ◽  
pp. 2017-2024
Author(s):  
Weiwei Zhang ◽  
Ziyi Li ◽  
Nana Wei ◽  
Hua-Jun Wu ◽  
Xiaoqi Zheng
Keyword(s):  
Real Data ◽  
R Package ◽  
Differential Methylation ◽  
Least Square ◽  
Epigenetic Mechanism ◽  
Supplementary Information ◽  
Cpg Sites ◽  
Tumor Purity ◽  
Different Sources ◽  
Normal Controls

Abstract Motivation Inference of differentially methylated (DM) CpG sites between two groups of tumor samples with different geno- or pheno-types is a critical step to uncover the epigenetic mechanism of tumorigenesis, and identify biomarkers for cancer subtyping. However, as a major source of confounding factor, uneven distributions of tumor purity between two groups of tumor samples will lead to biased discovery of DM sites if not properly accounted for. Results We here propose InfiniumDM, a generalized least square model to adjust tumor purity effect for differential methylation analysis. Our method is applicable to a variety of experimental designs including with or without normal controls, different sources of normal tissue contaminations. We compared our method with conventional methods including minfi, limma and limma corrected by tumor purity using simulated datasets. Our method shows significantly better performance at different levels of differential methylation thresholds, sample sizes, mean purity deviations and so on. We also applied the proposed method to breast cancer samples from TCGA database to further evaluate its performance. Overall, both simulation and real data analyses demonstrate favorable performance over existing methods serving similar purpose. Availability and implementation InfiniumDM is a part of R package InfiniumPurify, which is freely available from GitHub (https://github.com/Xiaoqizheng/InfiniumPurify). Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals DEsingle for detecting three types of differential expression in single-cell RNA-seq data

2017 ◽  
Author(s):  
Zhun Miao ◽  
Ke Deng ◽  
Xiaowo Wang ◽  
Xuegong Zhang
Keyword(s):  
Single Cell ◽  
Differential Expression ◽  
Negative Binomial ◽  
Single Cells ◽  
R Package ◽  
Supplementary Information ◽  
Binomial Model ◽  
Supplementary Data ◽  
Rna Seq ◽  
Real Zeros

AbstractSummaryThe excessive amount of zeros in single-cell RNA-seq data include “real” zeros due to the on-off nature of gene transcription in single cells and “dropout” zeros due to technical reasons. Existing differential expression (DE) analysis methods cannot distinguish these two types of zeros. We developed an R package DEsingle which employed Zero-Inflated Negative Binomial model to estimate the proportion of real and dropout zeros and to define and detect 3 types of DE genes in single-cell RNA-seq data with higher accuracy.Availability and ImplementationThe R package DEsingle is freely available at https://github.com/miaozhun/DEsingle and is under Bioconductor’s consideration [email protected] informationSupplementary data are available at bioRxiv online.

Sign in / Sign up

Export Citation Format

Share Document