Boolean implication analysis unveils candidate universal relationships in microbiome data

Abstract

Background Microbiomes consist of bacteria, viruses, and other microorganisms, and are responsible for many different functions in both organisms and the environment. Past analyses of microbiomes focused on using correlation to determine linear relationships between microbes and diseases. Weak correlations due to nonlinearity between microbe pairs may cause researchers to overlook critical components of the data. With the abundance of available microbiome, we need a method that comprehensively studies microbiomes and how they are related to each other. Results We collected publicly available datasets from human, environment, and animal samples to determine both symmetric and asymmetric Boolean implication relationships between a pair of microbes. We then found relationships that are potentially invariants, meaning they will hold in any microbe community. In other words, if we determine there is a relationship between two microbes, we expect the relationship to hold in almost all contexts. We discovered that around 330,000 pairs of microbes universally exhibit the same relationship in almost all the datasets we studied, thus making them good candidates for invariants. Our results also confirm known biological properties and seem promising in terms of disease diagnosis. Conclusions Since the relationships are likely universal, we expect them to hold in clinical settings, as well as general populations. If these strong invariants are present in disease settings, it may provide insight into prognostic, predictive, or therapeutic properties of clinically relevant diseases. For example, our results indicate that there is a difference in the microbe distributions between patients who have or do not have IBD, eczema and psoriasis. These new analyses may improve disease diagnosis and drug development in terms of accuracy and efficiency.

Detecting Boolean Asymmetric Relationships with a Loop Counting Technique and its Implications for Analyzing Heterogeneity within Gene Expression Datasets

“…Due to their importance and universality, BARs have been analyzed in many recent research efforts [6][7][8][9]. As just one example, [10] combined the use of a StepMiner algorithm and a sparsity test to find statistically-significant BARs, which can be generalized to find three-way-instead of merely pairwise-BARs as a way to reduce the computational burden and involve more general interactions between genes [11].…”

“…Along similar lines, significant BARs have been found associated with celldifferentiation, cellular-development and tissue-function [9], [34]–[37]. More recent work has shown that, while many BARs differentiate between different cell-types in the micro-biome, there are also many BARs that are strongly conserved across microorganisms, potentially highlighting ‘universal’ gene-functions important for survival [6].…”

“…Generally speaking, most common methods for finding/identifying genetic relationships focus on positive and negative correlations, which we'll refer to below as symmetric relationships [2][3][4]. In this paper, however, we'll focus on asymmetric relationships, which are a different kind of genetic relationship introduced and studied in [5][6][7][8][9].…”

“…Our main goal is to introduce a version of ‘biclustering’ – a form of heterogeneity analysis typically defined in terms of symmetric relationships – to also include these asymmetric relationships. Since there is growing evidence that a significant fraction of biologically-relevant gene-gene interactions are indeed asymmetric [6], our ultimate goal is to provide a simple preprocessing strategy that can be used to extend current biclustering tools to capture these asymmetric interactions.…”

Boolean implication analysis of single-cell data predicts retinal cell type markers

“…Our methods have not previously been applied to stem cell-derived retinal cell types, but have yielded insights into changes in transcriptional profiles of healthy retina and retinoblastoma [74]. The StepMiner and BooleanNet algorithms were developed for microarray data by Sahoo et al to identify Boolean implication relationships between genes, but have since been applied to a wide variety of high-throughput data, such as RNA-seq, scRNA-seq and microbiome data [68,[75][76][77].…”