Identifying Heritable Communities of Microbiome by Root-Unifrac and Wishart Distribution

06/06/2018
by   Yunfan Tang, et al.
0

We introduce a method to identify heritable microbiome communities when the input is a pairwise dissimilarity matrix among all samples. Current methods target each taxon individually and are unable to take advantage of their phylogenetic relationships. In contrast, our approach focuses on community heritability by using the root-Unifrac to summarize the microbiome samples through their pairwise dissimilarities while taking the phylogeny into account. The resulting dissimilarity matrix is then transformed into an outer product matrix and further modeled through a Wishart distribution with the same set of variance components as in the univariate model. Directly modeling the entire dissimilarity matrix allows us to bypass any dimension reduction steps. An important contribution of our work is to prove the positive definiteness of such outer product matrix, hence the applicability of the Wishart distribution. Simulation shows that this community heritability approach has higher power than existing methods to identify heritable groups of taxa. Empirical results on the TwinsUK dataset are also provided.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
06/03/2019

Transformed Central Quantile Subspace

We present a dimension reduction technique for the conditional quantiles...
research
07/09/2019

A method for computing the Perron-Frobenius root for primitive matrices

For a nonnegative matrix, the eigenvalue with the maximum magnitude or P...
research
04/14/2023

Subsampling-Based Modified Bayesian Information Criterion for Large-Scale Stochastic Block Models

Identifying the number of communities is a fundamental problem in commun...
research
06/28/2021

Variance Reduction for Matrix Computations with Applications to Gaussian Processes

In addition to recent developments in computing speed and memory, method...
research
08/12/2018

Outer Product-based Neural Collaborative Filtering

In this work, we contribute a new multi-layer neural network architectur...
research
06/02/2016

Generalized Root Models: Beyond Pairwise Graphical Models for Univariate Exponential Families

We present a novel k-way high-dimensional graphical model called the Gen...

Please sign up or login with your details

Forgot password? Click here to reset