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

2021 ◽  
Vol 12 ◽  
Author(s):  
Yuchen Zhang ◽  
Lina Zhu ◽  
Xin Wang

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.

2017 ◽  
Author(s):  
Yunan Zhu ◽  
Ivor Cribben

AbstractSparse graphical models are frequently used to explore both static and dynamic functional brain networks from neuroimaging data. However, the practical performance of the models has not been studied in detail for brain networks. In this work, we have two objectives. First, we compare several sparse graphical model estimation procedures and several selection criteria under various experimental settings, such as different dimensions, sample sizes, types of data, and sparsity levels of the true model structures. We discuss in detail the superiority and deficiency of each combination. Second, in the same simulation study, we show the impact of autocorrelation and whitening on the estimation of functional brain networks. We apply the methods to a resting-state functional magnetic resonance imaging (fMRI) data set. Our results show that the best sparse graphical model, in terms of detection of true connections and having few false-positive connections, is the smoothly clipped absolute deviation (SCAD) estimating method in combination with the Bayesian information criterion (BIC) and cross-validation (CV) selection method. In addition, the presence of autocorrelation in the data adversely affects the estimation of networks but can be helped by using the CV selection method. These results question the validity of a number of fMRI studies where inferior graphical model techniques have been used to estimate brain networks.


2018 ◽  
Vol 19 (5) ◽  
pp. 545-568 ◽  
Author(s):  
Geneviéve Robin ◽  
Christophe Ambroise ◽  
Stéphane Robin

Graphical network inference is used in many fields such as genomics or ecology to infer the conditional independence structure between variables, from measurements of gene expression or species abundances for instance. In many practical cases, not all variables involved in the network have been observed, and the samples are actually drawn from a distribution where some variables have been marginalized out. This challenges the sparsity assumption commonly made in graphical model inference, since marginalization yields locally dense structures, even when the original network is sparse. We present a procedure for inferring Gaussian graphical models when some variables are unobserved, that accounts both for the influence of missing variables and the low density of the original network. Our model is based on the aggregation of spanning trees, and the estimation procedure on the expectation-maximization algorithm. We treat the graph structure and the unobserved nodes as missing variables and compute posterior probabilities of edge appearance. To provide a complete methodology, we also propose several model selection criteria to estimate the number of missing nodes. A simulation study and an illustration on flow cytometry data reveal that our method has favourable edge detection properties compared to existing graph inference techniques. The methods are implemented in an R package.


2019 ◽  
Vol 25 (7) ◽  
pp. 750-773 ◽  
Author(s):  
Pabitra Narayan Samanta ◽  
Supratik Kar ◽  
Jerzy Leszczynski

The rapid advancement of computer architectures and development of mathematical algorithms offer a unique opportunity to leverage the simulation of macromolecular systems at physiologically relevant timescales. Herein, we discuss the impact of diverse structure-based and ligand-based molecular modeling techniques in designing potent and selective antagonists against each adenosine receptor (AR) subtype that constitutes multitude of drug targets. The efficiency and robustness of high-throughput empirical scoring function-based approaches for hit discovery and lead optimization in the AR family are assessed with the help of illustrative examples that have led to nanomolar to sub-micromolar inhibition activities. Recent progress in computer-aided drug discovery through homology modeling, quantitative structure-activity relation, pharmacophore models, and molecular docking coupled with more accurate free energy calculation methods are reported and critically analyzed within the framework of structure-based virtual screening of AR antagonists. Later, the potency and applicability of integrated molecular dynamics (MD) methods are addressed in the context of diligent inspection of intricated AR-antagonist binding processes. MD simulations are exposed to be competent for studying the role of the membrane as well as the receptor flexibility toward the precise evaluation of the biological activities of antagonistbound AR complexes such as ligand binding modes, inhibition affinity, and associated thermodynamic and kinetic parameters.


Biometrika ◽  
2020 ◽  
Author(s):  
S Na ◽  
M Kolar ◽  
O Koyejo

Abstract Differential graphical models are designed to represent the difference between the conditional dependence structures of two groups, thus are of particular interest for scientific investigation. Motivated by modern applications, this manuscript considers an extended setting where each group is generated by a latent variable Gaussian graphical model. Due to the existence of latent factors, the differential network is decomposed into sparse and low-rank components, both of which are symmetric indefinite matrices. We estimate these two components simultaneously using a two-stage procedure: (i) an initialization stage, which computes a simple, consistent estimator, and (ii) a convergence stage, implemented using a projected alternating gradient descent algorithm applied to a nonconvex objective, initialized using the output of the first stage. We prove that given the initialization, the estimator converges linearly with a nontrivial, minimax optimal statistical error. Experiments on synthetic and real data illustrate that the proposed nonconvex procedure outperforms existing methods.


EBioMedicine ◽  
2021 ◽  
Vol 70 ◽  
pp. 103525
Author(s):  
Abhijith Biji ◽  
Oyahida Khatun ◽  
Shachee Swaraj ◽  
Rohan Narayan ◽  
Raju S. Rajmani ◽  
...  

BMC Genomics ◽  
2021 ◽  
Vol 22 (1) ◽  
Author(s):  
Christos Dimitrakopoulos ◽  
Sravanth Kumar Hindupur ◽  
Marco Colombi ◽  
Dritan Liko ◽  
Charlotte K. Y. Ng ◽  
...  

Abstract Background Genetic aberrations in hepatocellular carcinoma (HCC) are well known, but the functional consequences of such aberrations remain poorly understood. Results Here, we explored the effect of defined genetic changes on the transcriptome, proteome and phosphoproteome in twelve tumors from an mTOR-driven hepatocellular carcinoma mouse model. Using Network-based Integration of multi-omiCS data (NetICS), we detected 74 ‘mediators’ that relay via molecular interactions the effects of genetic and miRNA expression changes. The detected mediators account for the effects of oncogenic mTOR signaling on the transcriptome, proteome and phosphoproteome. We confirmed the dysregulation of the mediators YAP1, GRB2, SIRT1, HDAC4 and LIS1 in human HCC. Conclusions This study suggests that targeting pathways such as YAP1 or GRB2 signaling and pathways regulating global histone acetylation could be beneficial in treating HCC with hyperactive mTOR signaling.


2020 ◽  
Vol 992 ◽  
pp. 658-662
Author(s):  
M.A. Mokeev ◽  
L.A. Urkhanova ◽  
A.N. Khagleev ◽  
Denis B. Solovev

Mechanical, chemical and plasma treatment are the main kind of treatment of polytetrafluoroethylene (PTFE) films. Each method is different from each other by the adhesive force: the value of the wetting angle. Mechanical treatment allows different particles to permeate into the structure of the polymer. Chemical treatment creates new functional groups on the polymer surface, but this method is toxic and dangerous. Plasma treatment, in a glow discharge non-thermal plasma, is a more ecological and practical method. The experiment showed that the plasma treatment successfully increases the adhesion, this has been proven by infrared spectroscopy and scanning electron microscopy. According to the obtained data of the wetting angle, the regression equation was derived. A graphical model is constructed by regression equations allows you to determine the main processing factor and choose the optimal values of treatment.


Cancers ◽  
2019 ◽  
Vol 11 (7) ◽  
pp. 983 ◽  
Author(s):  
Otília Menyhart ◽  
Tatsuhiko Kakisaka ◽  
Lőrinc Sándor Pongor ◽  
Hiroyuki Uetake ◽  
Ajay Goel ◽  
...  

Background: Numerous driver mutations have been identified in colorectal cancer (CRC), but their relevance to the development of targeted therapies remains elusive. The secondary effects of pathogenic driver mutations on downstream signaling pathways offer a potential approach for the identification of therapeutic targets. We aimed to identify differentially expressed genes as potential drug targets linked to driver mutations. Methods: Somatic mutations and the gene expression data of 582 CRC patients were utilized, incorporating the mutational status of 39,916 and the expression levels of 20,500 genes. To uncover candidate targets, the expression levels of various genes in wild-type and mutant cases for the most frequent disruptive mutations were compared with a Mann–Whitney test. A survival analysis was performed in 2100 patients with transcriptomic gene expression data. Up-regulated genes associated with worse survival were filtered for potentially actionable targets. The most significant hits were validated in an independent set of 171 CRC patients. Results: Altogether, 426 disruptive mutation-associated upregulated genes were identified. Among these, 95 were linked to worse recurrence-free survival (RFS). Based on the druggability filter, 37 potentially actionable targets were revealed. We selected seven genes and validated their expression in 171 patient specimens. The best independently validated combinations were DUSP4 (p = 2.6 × 10−12) in ACVR2A mutated (7.7%) patients; BMP4 (p = 1.6 × 10−04) in SOX9 mutated (8.1%) patients; TRIB2 (p = 1.35 × 10−14) in ACVR2A mutated patients; VSIG4 (p = 2.6 × 10−05) in ANK3 mutated (7.6%) patients, and DUSP4 (p = 7.1 × 10−04) in AMER1 mutated (8.2%) patients. Conclusions: The results uncovered potentially druggable genes in colorectal cancer. The identified mutations could enable future patient stratification for targeted therapy.


2019 ◽  
Author(s):  
Linus Holm ◽  
Gustaf Wadenholt ◽  
Paul Schrater

Humans often appear to desire information for its own sake, but it is presently unclear what drives this desire. The important role that resolving uncertainty plays in stimulating information seeking has suggested a tight coupling between the intrinsic motivation to gather information and performance gains, and has been construed as a drive for long-term learning. Using a simple asteroid-avoidance game that allows us to study learning and information seeking at an experimental time-scale, we show that we can separate the incentive for information-seeking from a long-term learning outcome, and show that information-seeking is best predicted by per-trial outcome uncertainty. Specifically, our 43 participants were more willing to take time penalties for feedback on trials with uncertain outcomes. We found strong group (R2 = .97) and individual level (mean R2 = .44) support for a linear relationship between feedback request rate and information gain as determined by per-trial outcome uncertainty. This information better reflects filling in the gaps of the episodic record of choice outcomes than long-term skill acquisition or assessment. Our results suggest that this easy to compute quantity can drive information-seeking, potentially allowing simple organisms to intelligently gather information without having to anticipate the impact on future performance.


2021 ◽  
Author(s):  
Félix Raimundo ◽  
Laetitia Papaxanthos ◽  
Céline Vallot ◽  
Jean-Philippe Vert

AbstractSingle-cell omics technologies produce large quantities of data describing the genomic, transcriptomic or epigenomic profiles of many individual cells in parallel. In order to infer biological knowledge and develop predictive models from these data, machine learning (ML)-based model are increasingly used due to their flexibility, scalability, and impressive success in other fields. In recent years, we have seen a surge of new ML-based method development for low-dimensional representations of single-cell omics data, batch normalization, cell type classification, trajectory inference, gene regulatory network inference or multimodal data integration. To help readers navigate this fast-moving literature, we survey in this review recent advances in ML approaches developed to analyze single-cell omics data, focusing mainly on peer-reviewed publications published in the last two years (2019-2020).


Sign in / Sign up

Export Citation Format

Share Document