Ilya Shmulevich | bioRxiv Science

bioRxiv2016

Solving the influence maximization problem reveals regulatory organization of the yeast cell cycle.

The Influence Maximization Problem (IMP) aims to discover the set of nodes with the greatest influence on network dynamics. The problem has previously been applied in epidemiology and social network analysis. Here, we demonstrate the application to cell cycle regulatory network analysis of Saccharomyces cerevisiae.\n\nFundamentally, gene regulation is linked to the flow of information. Therefore, our implementation of the IMP was framed as an information theoretic problem on a diffusion network. Utilizing all regulatory edges from YeastMine, gene expression dynamics were encoded as edge weights using a variant of time lagged transfer entropy, a method for quantifying information transfer between variables. Influence, for a particular number of sources, was measured using a diffusion model based on Markov chains with absorbing states. By maximizing over different numbers of sources, an influence ranking on genes was produced.\n\nThe influence ranking was compared to other metrics of network centrality. Although top genes from each centrality ranking contained well-known cell cycle regulators, there was little agreement and no clear winner. However, it was found that influential genes tend to directly regulate or sit upstream of genes ranked by other centrality measures. This is quantified by computing node reachability between gene sets; on average, 59% of central genes can be reached when starting from the influential set, compared to 7% of influential genes when starting at another centrality measure.\n\nThe influential nodes act as critical sources of information flow, potentially having a large impact on the state of the network. Biological events that affect influential nodes and thereby affect information flow could have a strong effect on network dynamics, potentially leading to disease.\n\nCode and example data can be found at: https://github.com/Gibbsdavidl/miergolf\n\nAuthor SummaryThe Influence Maximization Problem (IMP) is general and is applied in fields such as epidemiology, social network analysis, and as shown here, biological network analysis. The aim is to discover the set of regulatory genes with the greatest influence in the network dynamics. As gene regulation, fundamentally, is about the flow of information, the IMP was framed as an information theoretic problem. Dynamics were encoded as edge weights using time lagged transfer entropy, a quantity that defines information transfer across variables. The information flow was accomplished using a diffusion model based on Markov chains with absorbing states. Ant optimization was applied to solve the subset selection problem, recovering the most influential nodes.The influential nodes act as critical sources of information flow, potentially affecting the network state. Biological events that impact the influential nodes and thereby affecting normal information flow, could have a strong effect on the network, potentially leading to disease.

Bioinformatics↗

bioRxiv2016

Logic models to predict continuous outputs based on binary inputs with an application to personalized cancer therapy

Mining large datasets using machine learning approaches often leads to models that are hard to interpret and not amenable to the generation of hypotheses that can be experimentally tested. Finding actionable knowledge is becoming more important, but also more challenging as datasets grow in size and complexity. We present Logic Optimization for Binary Input to Continuous Output (LOBICO), a computational approach that infers small and easily interpretable logic models of binary input features that explain a binarized continuous output variable. Although the continuous output variable is binarized prior to optimization, the continuous information is retained to find the optimal logic model. Applying LOBICO to a large cancer cell line panel, we find that logic combinations of multiple mutations are more predictive of drug response than single gene predictors. Importantly, we show that the use of the continuous information leads to robust and more accurate logic models. LOBICO is formulated as an integer programming problem, which enables rapid computation on large datasets. Moreover, LOBICO implements the ability to uncover logic models around predefined operating points in terms of sensitivity and specificity. As such, it represents an important step towards practical application of interpretable logic models.

Bioinformatics↗

bioRxiv2015

Combining Dependent P-values with an Empirical Adaptation of Brown’s Method

Motivation: Combining P-values from multiple statistical tests is a common exercise in bioinformatics. However, this procedure is non-trivial for dependent P-values. Here we discuss an empirical adaptation of Browns Method (an extension of Fishers Method) for combining dependent P-values which is appropriate for the correlated data sets found in high-throughput biological experiments.\n\nResults: We show that Fishers Method is biased when used on dependent sets of P-values with both simulated data and gene expression data from The Cancer Genome Atlas (TCGA). When applied on the same data sets, the Empirical Browns Method provides a better null distribution and a more conservative result.\n\nAvailability: The Empirical Browns Method is available in Python, R, and MATLAB and can be obtained from https://github.com/IlyaLab/CombiningDependentPvaluesUsingEBM.1

Bioinformatics↗