NEM-Tar: A Probabilistic Graphical Model for Cancer Regulatory Network Inference and Prioritization of Potential Therapeutic Targets From Multi-Omics Data

Targeted therapy has been widely adopted as an effective treatment strategy to battle against cancer. However, cancers are not single disease entities, but comprising multiple molecularly distinct subtypes, and the heterogeneity nature prevents precise selection of patients for optimized therapy. Dissecting cancer subtype-specific signaling pathways is crucial to pinpointing dysregulated genes for the prioritization of novel therapeutic targets. Nested effects models (NEMs) are a group of graphical models that encode subset relations between observed downstream effects under perturbations to upstream signaling genes, providing a prototype for mapping the inner workings of the cell. In this study, we developed NEM-Tar, which extends the original NEMs to predict drug targets by incorporating causal information of (epi)genetic aberrations for signaling pathway inference. An information theory-based score, weighted information gain (WIG), was proposed to assess the impact of signaling genes on a specific downstream biological process of interest. Subsequently, we conducted simulation studies to compare three inference methods and found that the greedy hill-climbing algorithm demonstrated the highest accuracy and robustness to noise. Furthermore, two case studies were conducted using multi-omics data for colorectal cancer (CRC) and gastric cancer (GC) in the TCGA database. Using NEM-Tar, we inferred signaling networks driving the poor-prognosis subtypes of CRC and GC, respectively. Our model prioritized not only potential individual drug targets such as HER2, for which FDA-approved inhibitors are available but also the combinations of multiple targets potentially useful for the design of combination therapies.

Metabolic pathway prediction using non-negative matrix factorization with improved precision

Machine learning provides a probabilistic framework for metabolic pathway inference from genomic sequence information at different levels of complexity and completion. However, several challenges including pathway features engineering, multiple mapping of enzymatic reactions and emergent or distributed metabolism within populations or communities of cells can limit prediction performance. In this paper, we present triUMPF, triple non-negative matrix factorization (NMF) with community detection for metabolic pathway inference, that combines three stages of NMF to capture myriad relationships between enzymes and pathways within a graph network. This is followed by community detection to extract higher order structure based on the clustering of vertices which share similar statistical properties. We evaluated triUMPF performance using experimental datasets manifesting diverse multi-label properties, including Tier 1 genomes from the BioCyc collection of organismal Pathway/Genome Databases and low complexity microbial communities. Resulting performance metrics equaled or exceeded other prediction methods on organismal genomes with improved precision on multi-organismal datasets.

Degradation of acetaminophen (ACT) by ozone/persulfate oxidation process: experimental and degradation pathways

Purpose The purpose of this study to investigate acetaminophen (ACT) degradation efficiencies by using ozone/persulfate oxidation process in a batch reactor. In addition, the effects of various parameters on the ACT removal efficiency toward pathway inference of ACT degradation were investigated. Design/methodology/approach The experiments were in the 2 L glass vessels. Ozone gas with flow rate at 70 L.h−1 was produced by ozone generator. After the adjustment of the pH, various dosages of persulfate (1, 3, 5, 7 and 9 mmol.L−1) were then added to the 500 mL ACT-containing solution with 150 mg.L−1 of concentration. Afterward, ozone gas was diffused in glass vessels. The solution after reaction flowed into the storage tank for the detection. The investigated parameters included pH and the amount of ozone and persulfate addition. For comparison of the ACT degradation efficiency, ozone/persulfate, ozone and persulfate oxidation in reactor was carried out. The ACT concentration using a HPLC system equipped with 2998 PDA detector was determined at an absorbance of 242 nm. Findings ACT degradation percentage by using ozone or persulfate in the process were at 63.7% and 22.3%, respectively, whereas O3/persulfate oxidation process achieved degradation percentage at 91.4% in 30 min. Degradation efficiency of ACT was affected by different parameter like pH and addition of ozone or persulfate, and highest degradation obtained when pH and concentrations of persulfate and ozone was 10 and 3 mmol.L−1 and 60 mg.L−1, respectively. O3, OH• and SO4− were evidenced to be the radicals for degradation of ACT through direct and indirect oxidation. Gas chromatography–mass spectrometer analysis showed intermediates including N-(3,4-dihydroxyphenyl) formamide, hydroquinone, benzoic acid, 4-methylbenzene-1,2-diol, 4-aminophenol. Practical implications This study provided a simple and effective way for degradation of activated ACT as emerging contaminants from aqueous solution. This way was conducted to protect environment from one of the most important and abundant pharmaceutical and personal care product in aquatic environments. Originality/value There are two main innovations. One is that the novel process is performed successfully for pharmaceutical degradation. The other is that the optimized conditions are obtained. In addition, the effects of various parameters on the ACT removal efficiency toward pathway inference of ACT degradation were investigated.

Pathway mining in functional genomics: An integrative approach to delineate boolean relationships between Src and its targets

Motivation: There are important molecular information hidden in the ocean of big data could be achieved by recognizing true relationships between different molecules. Human mind is very limited to find all molecular connections. Therefore, we introduced an integrated data mining strategy to find all possible relationships between molecular components in a biological context. To demonstrate how this approach works, we applied it on proto-oncogene c-Src. Results: Here we applied a data mining scheme on genomic, literature and signaling databases to obtain necessary biological information for pathway inference. Using R programming language, two large edgelists were constructed from KEGG and OmniPath signaling databases. Next, An R script was developed by which pathways were discovered by assembly of edge information in the constructed signaling networks. Then, valid pathways were distinguished from the invalid ones using molecular information in articles and genomic data analysis. Pathway inference was performed on predicted pathways starting with Src and ending with the DEGs whose expression were affected by c-Src overactivation. Moreover, some positive and negative feedback loops were proposed based on the gene expression results. In fact, this simple but practical flowchart will open new insights into interactions between cellular components and help biologists look for new possible molecular relationships that have not been reported neither in signaling databases nor as a signaling pathway.

Mechanisms for the adverse effects of late gestational increases in maternal cortisol on the heart revealed by transcriptomic analyses of the fetal septum

We have previously shown in sheep that 10 days of modest chronic increase in maternal cortisol resulting from maternal infusion of cortisol (1 mg/kg/day) caused fetal heart enlargement and Purkinje cell apoptosis. In subsequent studies we extended the cortisol infusion to term, finding a dramatic incidence of stillbirth in the pregnancies with chronically increased cortisol. To investigate effects of maternal cortisol on the heart, we performed transcriptomic analyses on the septa using ovine microarrays and Webgestalt and Cytoscape programs for pathway inference. Analyses of the transcriptomic effects of maternal cortisol infusion for 10 days (130 day cortisol vs 130 day control), or ∼25 days (140 day cortisol vs 140 day control) and of normal maturation (140 day control vs 130 day control) were performed. Gene ontology terms related to immune function and cytokine actions were significantly overrepresented as genes altered by both cortisol and maturation in the septa. After 10 days of cortisol, growth factor and muscle cell apoptosis pathways were significantly overrepresented, consistent with our previous histologic findings. In the term fetuses (∼25 days of cortisol) nutrient pathways were significantly overrepresented, consistent with altered metabolism and reduced mitochondria. Analysis of mitochondrial number by mitochondrial DNA expression confirmed a significant decrease in mitochondria. The metabolic pathways modeled as altered by cortisol treatment to term were different from those modeled during maturation of the heart to term, and thus changes in gene expression in these metabolic pathways may be indicative of the fetal heart pathophysiologies seen in pregnancies complicated by stillbirth, including gestational diabetes, Cushing's disease and chronic stress.

Learning Dysregulated Pathways in Cancers from Differential Variability Analysis

Analysis of gene sets can implicate activity in signaling pathways that is responsible for cancer initiation and progression, but is not discernible from the analysis of individual genes. Multiple methods and software packages have been developed to infer pathway activity from expression measurements for set of genes targeted by that pathway. Broadly, three major methodologies have been proposed: over-representation, enrichment, and differential variability. Both over-representation and enrichment analyses are effective techniques to infer differentially regulated pathways from gene sets with relatively consistent differentially expressed (DE) genes. Specifically, these algorithms aggregate statistics from each gene in the pathway. However, they overlook multivariate patterns related to gene interactions and variations in expression. Therefore, the analysis of differential variability of multigene expression patterns can be essential to pathway inference in cancers. The corresponding methodologies and software packages for such multivariate variability analysis of pathways are reviewed here. We also introduce a new, computationally efficient algorithm, expression variation analysis (EVA), which has been implemented along with a previously proposed algorithm, Differential Rank Conservation (DIRAC), in an open source R package, gene set regulation (GSReg). EVA inferred similar pathways as DIRAC at reduced computational costs. Moreover, EVA also inferred different dysregulated pathways than those identified by enrichment analysis.