Statistical Methods for Detecting Differentially Abundant Features in Clinical Metagenomic Samples

By James Robert White, Niranjan Nagarajan, Mihai Pop. PLoS Comput. Biol., 2009

1,517 citations recorded by the source corpus.

Numerous studies are currently underway to characterize the microbial communities inhabiting our world. These studies aim to dramatically expand our understanding of the microbial biosphere and, more importantly, hope to reveal the secrets of the complex symbiotic relationship between us and our commensal bacterial microflora. An important prerequisite for such discoveries are computational tools that are able to rapidly and accurately compare large datasets generated from complex bacterial communities to identify features that distinguish them. We present a statistical method for comparing clinical metagenomic samples from two treatment populations on the basis of count data (e.g. as obtained through sequencing) to detect differentially abundant features. Our method, Metastats, employs the false discovery rate to improve specificity in high-complexity environments, and separately handles sparsely-sampled features using Fisher's exact test. Under a variety of simulations, we show that Metastats performs well compared to previously used methods, and significantly outperforms other methods for features with sparse counts. We demonstrate the utility of our method on several datasets including a 16S rRNA survey of obese and lean human gut microbiomes, COG functional profiles of infant and mature gut microbiomes, and bacterial and viral metabolic subsystem data inferred from random sequencing of 85 metagenomes. The application of our method to the obesity dataset reveals differences between obese and lean subjects not reported in the original study. For the COG and subsystem datasets, we provide the first statistically rigorous assessment of the differences between these populations. The methods described in this paper are the first to address clinical metagenomic datasets comprising samples from multiple subjects. Our methods are robust across datasets of varied complexity and sampling level. While designed for metagenomic applications, our software can also be applied to digital gene expression studies (e.g. SAGE). A web server implementation of our methods and freely available source code can be found at http://metastats.cbcb.umd.edu/.

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/a80ba5b7c4ea6551a6f4ab64fde15819125bd1cb

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Defining the Human Microbiome (2012, methodology): “Another powerful tool is the use of machine learning and statistical techniques to build predictive models of taxa[25] or functions[26] that discriminate between groups of samples.”
Hypothesis Testing and Power Calculations for Taxonomic-Based Human Microbiome Data (2012, methodology): “…these methods require a significant reduction of information, such as Unifrac [7] which reduces sequence data to pairwise distances, or ignoring correlations and the multivariate structure inherent in microbiome data, such as Metastats [12] which does univariate ‘one-taxa-at-a-time’ analyses.”
The microbiota modulates gut physiology and behavioral abnormalities associated with autism (2013, positive): “Identification of Differences in Specific OTUs Key OTUs were identified using: (1) Metastats comparison (White et al., 2009) and (2) Random Forests algorithm, under QIIME (Knights et al.”
Computational meta'omics for microbial community studies (2013, background): “It remains an area of active research, with current options including categorical nonparametric biomarker discovery (White et al , 2009; Parks and Beiko, 2010; Segata et al , 2011) and appropriately transformed regression models (Chen et al , 2012).”
METAGENassist: a comprehensive web server for comparative metagenomics (2012, positive): “Mothur has recently added the Metastats tool (22) for determining which taxa are differentially abundant between two groups of samples.”
Metagenomic systems biology of the human gut microbiome reveals topological shifts associated with obesity and inflammatory bowel disease (2011, background): “In contrast, metagenomic shotgun sequencing of community DNA and a gene-centric comparative approach (8, 13, 14) may capture functional differences in the metabolic potential of the community.”
Metagenomic biomarker discovery and explanation (2011, methodology): “Metastats [45] seems to achieve results very similar to KW (Additional file 7) with the same disadvantages with respect to LEfSe.”
Machine Learning Meta-analysis of Large Metagenomic Datasets: Tools and Biological Insights (2016, background): “11 strictly statistical assessments [46] or statistical assessment coupled with effect size [47].”
Common and distinguishing features of the bacterial and fungal communities in biological soil crusts and shrub root zone soils (2014, methodology): “Identification of differentially abundant features in the bacterial and fungal rRNA gene libraries and shotgun metagenomes was performed using White’s non-parametric t-test using the Bonferroni correction for multiple comparisons (White et al., 2009) as implemented in the STAMP software package (Par”
The Cirrhosis Dysbiosis Ratio defines Changes in the Gut Microbiome Associated with Cirrhosis and its Complications (2013, positive): “Similar to the other outcomes, patients who developed ACLF had higher endotoxin (1.6 vs. 1.0, p = 0.04) and lower CDR (0.6 vs. 1.3, p = 0.01) due to a significantly lower abundance of gram-positives on Metastats; Clostridiales XIV (0 vs. 2.3”