scholarly journals Mining signaling flow to interpret mechanisms of synergy of drug combinations using deep graph neural networks

2021 ◽  
Author(s):  
Heming Zhang ◽  
Yixin Chen ◽  
Philip R Payne ◽  
Fuhai Li
Keyword(s):  
Drug Resistance ◽  
Signaling Pathways ◽  
Drug Combination ◽  
Drug Targets ◽  
Clinical Decision Making ◽  
Clinical Decision ◽  
Signaling Network ◽  
Drug Combinations ◽  
Convolutional Network ◽  
Cancer Signaling

Complex signaling pathways/networks are believed to be responsible for drug resistance in cancer therapy. Drug combinations inhibiting multiple signaling targets within cancer-related signaling networks have the potential to reduce drug resistance. Deep learning models have been reported to predict drug combinations. However, these models are hard to be interpreted in terms of mechanism of synergy (MoS), and thus cannot well support the human-AI based clinical decision making. Herein, we proposed a novel computational model, DeepSignalingFlow, which seeks to address the preceding two challenges. Specifically, a graph convolutional network (GCN) was developed based on a core cancer signaling network consisting of 1584 genes, with gene expression and copy number data derived from 46 core cancer signaling pathways. The novel up-stream signaling-flow (from up-stream signaling to drug targets), and the down-stream signaling-flow (from drug targets to down-stream signaling), were designed using trainable weights of network edges. The numerical features (accumulated information due to the signaling-flows of the signaling network) of drug nodes that link to drug targets were then used to predict the synergy scores of such drug combinations. The model was evaluated using the NCI ALMANAC drug combination screening data. The evaluation results showed that the proposed DeepSignalingFlow model can not only predict drug combination synergy score, but also interpret potentially interpretable MoS of drug combinations.

scholarly journals Predicting effective drug combinations using gradient tree boosting based on features extracted from drug-protein heterogeneous network

2019 ◽  
Vol 20 (1) ◽  
Author(s):  
Hui Liu ◽  
Wenhao Zhang ◽  
Lixia Nie ◽  
Xiancheng Ding ◽  
Judong Luo ◽  
...  
Keyword(s):  
Drug Combination ◽  
Heterogeneous Network ◽  
Drug Targets ◽  
Complex Disease ◽  
Drug Combinations ◽  
Activity Level ◽  
Network Pharmacology ◽  
Data Set ◽  
Similarity Network ◽  
Heterogenous Network

Abstract Background Although targeted drugs have contributed to impressive advances in the treatment of cancer patients, their clinical benefits on tumor therapies are greatly limited due to intrinsic and acquired resistance of cancer cells against such drugs. Drug combinations synergistically interfere with protein networks to inhibit the activity level of carcinogenic genes more effectively, and therefore play an increasingly important role in the treatment of complex disease. Results In this paper, we combined the drug similarity network, protein similarity network and known drug-protein associations into a drug-protein heterogenous network. Next, we ran random walk with restart (RWR) on the heterogenous network using the combinatorial drug targets as the initial probability, and obtained the converged probability distribution as the feature vector of each drug combination. Taking these feature vectors as input, we trained a gradient tree boosting (GTB) classifier to predict new drug combinations. We conducted performance evaluation on the widely used drug combination data set derived from the DCDB database. The experimental results show that our method outperforms seven typical classifiers and traditional boosting algorithms. Conclusions The heterogeneous network-derived features introduced in our method are more informative and enriching compared to the primary ontology features, which results in better performance. In addition, from the perspective of network pharmacology, our method effectively exploits the topological attributes and interactions of drug targets in the overall biological network, which proves to be a systematic and reliable approach for drug discovery.

scholarly journals Deep convolutional approaches for the analysis of Covid-19 using chest X-Ray images from portable devices

2020 ◽  
Author(s):  
Joaquim de Moura ◽  
Lucía Ramos ◽  
Plácido L. Vidal ◽  
Milena Cruz ◽  
Laura Abelairas ◽  
...  
Keyword(s):  
Network Architecture ◽  
Clinical Decision Making ◽  
Clinical Decision ◽  
World Health ◽  
Portable Devices ◽  
Learning Approaches ◽  
Portable Equipment ◽  
Convolutional Network ◽  
X Ray ◽  
Chest X Ray

The recent human coronavirus disease (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been declared as a global pandemic on 11 March 2020 by the World Health Organization. Given the effects of COVID-19 in pulmonary tissues, chest radiography imaging plays an important role for the screening, early detection and monitoring of the suspected individuals. Hence, as the pandemic of COVID-19 progresses, there will be a greater reliance on the use of portable equipment for the acquisition of chest X-Ray images due to its accessibility, widespread availability and benefits regarding to infection control issues, minimizing the risk of cross contamination. This work presents novel fully automatic approaches specifically tailored for the classification of chest X-Ray images acquired by portable equipment into 3 different clinical categories: normal, pathological and COVID-19. For this purpose, two complementary deep learning approaches based on a densely convolutional network architecture are herein presented. The joint response of both approaches allows to enhance the differentiation between patients infected with COVID-19, patients with other diseases that manifest characteristics similar to COVID-19 and normal cases. The proposed approaches were validated over a dataset provided by the Radiology Service of the Complexo Hospitalario Universitario A Coruña (CHUAC) specifically retrieved for this research. Despite the poor quality of chest X-Ray images that is inherent to the nature of the portable equipment, the proposed approaches provided satisfactory results, allowing a reliable analysis of portable radiographs, to support the clinical decision-making process.

Novel mutational and pathway signatures in relapsed/refractory colorectal cancer patients.

2016 ◽  
Vol 34 (4_suppl) ◽  
pp. 591-591
Author(s):  
Benny Johnson ◽  
Laurence Cooke ◽  
Daruka Mahadevan
Keyword(s):  
Colorectal Cancer ◽  
Signaling Pathways ◽  
Clinical Decision Making ◽  
Gene Mutations ◽  
Clinical Decision ◽  
Braf V600e ◽  
Driver Mutations ◽  
Common Mutation ◽  
Mutational Signatures ◽  
Pathway Networks

591 Background: In the management of metastatic colorectal cancer (mCRC), all RAS (KRAS, NRAS) and BRAF V600E mutation status guides therapeutic options and identify a unique cohort of patients (pts) with a more aggressive clinical course. We hypothesized that relapsed/refractory CRC pts develop unique mutational signatures that guide standard and targeted therapy but also predict for therapeutic response, identify novel driver mutations and highlight key signaling pathways for clinical decision making. Methods: Relapsed/refractory mCRC pts (N=31) were molecularly profiled by NGS Caris Molecular Intelligence (IHC, FISH/CISH, NGS) and/or Foundation One (NGS, copy number). Samples were annotated by histology, primary and/or metastatic site, biopsy location, gene mutation, domain, topology, and mutation count. Web-based bioinformatics tools (Enrichr/IntAct) were utilized to analyze mutational profiles, identifying pathway-networks. Results: Pts included progressed on fluoropyrimidines, oxaliplatin, irinotecan, bevacizumab, cetuximab or panitumumab. Most common histology was adenocarcinoma followed by squamous cell carcinoma (colon N=28; rectal N=3). TP53 was the most common mutation followed by APC, KRAS, PIK3CA, BRAF, SMAD4, SPTA1, FAT1, PDGFRA, ATM, ROS1, ALK, CDKN2A, FBXW7, TGFBR2, NOTCH1 and HER3. Pts had on average >5 unique gene mutations. High mutational burden was not predictive for PD-1 (5 pts) or PD-L1 (1 pt) positivity. The most common activated signaling pathways were: ERRB2/HER2, FGFR, p38 activation through BRAF-MEK cascade via RIT and RIN, ARMS-mediated activation of MAPK cascade, and VEGFR2. Conclusions: Dominant oncogene mutations do not always equate with oncogenic dependence, therefore understanding the pathologic interactome in each patient is important to both identification of clinically relevant targets and choosing the next best therapy. Mutational signatures derived from corresponding ‘pathway-networks’ represent a meaningful tool to 1). Evaluate functional investigation in the laboratory, 2). Predict response to drug therapy, and 3). Guide rational drug combinations in relapsed/refractory mCRC pts entering targeted and immune checkpoint trials.

ABCB1 gene polymorphism associated with clinical factors can predict drug-resistant tuberculosis

2017 ◽  
Vol 131 (15) ◽  
pp. 1831-1840 ◽  
Author(s):  
Yasmin Pontual ◽  
Vanessa S.S. Pacheco ◽  
Sérgio P. Monteiro ◽  
Marcel S.B. Quintana ◽  
Marli J.M. Costa ◽  
...  
Keyword(s):  
Drug Resistance ◽  
Skin Color ◽  
Clinical Decision Making ◽  
Efflux Pump ◽  
Direct Sequencing ◽  
Previous Treatment ◽  
Clinical Decision ◽  
Control Group ◽  
Case Group ◽  
Abcb1 Gene

Polymorphism in the ABCB1 gene encoding P-glycoprotein, a transmembrane drug efflux pump, contributes to drug resistance and has been widely studied. However, their association with rifampicin and ethambutol resistance in tuberculosis (TB) patients is still unclear. Genotype/allele/haplotype frequencies in c.1236C > T (rs1128503), c.2677G > T/A (rs2032582), and c.3435C > T (rs1045642) were obtained from 218 patients. Of these, 80 patients with rifampicin and/or ethambutol resistance were selected as the case group and 138 patients were selected for the control group through the results of their culture and drug-sensitive tests. Patients aged <18 years and HIV-positive serologic tests were excluded. ABCB1 polymorphisms were determined using a PCR direct-sequencing approach, and restriction fragment length polymorphism (RFLP). A nomogram was constructed to simulate a combined prediction of the probability of anti-TB drug resistance, with factors including genotype c.1236C > T (rs1128503) (P=0.02), clinical form (P=0.03), previous treatment (P=0.01), and skin color (P=0.03), contributing up to 90% chance of developing anti-TB drug resistance. Considering genotype analyses, CT (rs1128503) demonstrated an increased chance of anti-TB drug resistance (odds ratio (OR): 2.34, P=0.02), while the analyses for ethambutol resistance revealed an association with a rare A allele (rs2032582) (OR: 12.91, P=0.01), the haplotype TTC (OR: 5.83, P=0.05), and any haplotype containing the rare A allele (OR: 7.17, P=0.04). ABCB1 gene polymorphisms in association with others risk factors contribute to anti-TB drug resistance, mainly ethambutol. The use of the nomogram described in the present study could contribute to clinical decision-making prior to starting TB treatment.

scholarly journals GraphSynergy: Network Inspired Deep Learning Model for Anti–Cancer Drug Combination Prediction

2021 ◽  
Author(s):  
Jiannan Yang ◽  
Zhongzhi Xu ◽  
William Wu ◽  
Qian Chu ◽  
Qingpeng Zhang
Keyword(s):  
Deep Learning ◽  
Cell Lines ◽  
Cancer Cell ◽  
Drug Combination ◽  
Drug Combinations ◽  
Cancer Drug ◽  
Ppi Network ◽  
Convolutional Network ◽  
Anti Cancer ◽  
Protein Modules

Abstract Compared with monotherapy, anti-cancer drug combination can provide effective therapy with less toxicity in cancer treatment. Recent studies found that the topological positions of protein modules related to the drugs and the cancer cell lines in the protein-protein interaction (PPI) network may reveal the effects of drugs. However, due to the size of the combinatorial space, identifying synergistic combinations of drugs from PPI network is computationally difficult. To address this challenge, we propose an end-to-end deep learning framework, namely Graph Convolutional Network for Drug Synergy (GraphSynergy), to make synergistic drug combination predictions. GraphSynergy adapts a spatial-based Graph Convolutional Network component to encode the high-order structure information of protein modules targeted by a pair of drugs, as well as the protein modules associated with a specific cancer cell line in the PPI network. The pharmacological effects of drug combinations are explicitly evaluated by their therapy and toxic scores. By introducing an attention component to automatically allocate contribution weights to the proteins, we show the ability of GraphSynergy to capture the pivotal proteins that play a part in both PPI network and biomolecular interactions between drug combinations and cancer cell lines. Experiments on two latest drug combination datasets demonstrate that GraphSynergy outperforms the state-of-the-art in predicting synergistic drug combinations. This study sheds light on using machine learning to discover effective combination therapies for cancer and other complex diseases.

Diffusion mapping of drug targets on disease signaling network elements reveals drug combination strategies

2017 ◽  
Author(s):  
Jielin Xu ◽  
Kelly Regan-Fendt ◽  
Siyuan Deng ◽  
William E. Carson ◽  
Philip R.O. Payne ◽  
...  
Keyword(s):  
Drug Combination ◽  
Drug Targets ◽  
Signaling Network ◽  
Combination Strategies

Chemoresistance in Ovarian Cancer: Prospects for New Drugs

2020 ◽  
Vol 20 ◽  
Author(s):  
Shivani Tendulkar ◽  
Suneel Dodamani
Keyword(s):  
Ovarian Cancer ◽  
Drug Resistance ◽  
Signaling Pathways ◽  
Drug Targets ◽  
Cellular Level ◽  
New Drugs ◽  
Therapeutic Outcomes ◽  
Pathway Activation ◽  
Efflux Pathway ◽  
Clinical Trails

: This review focuses on conventional treatment overview, signaling pathways and various reasons for drug resistance with understanding novel methods that can lead to effective therapies. Ovarian cancer is amongst the most common gynecological and lethal cancers in women from the age of 20-60. The survival rate is limited to 5 years due to diagnosis in subsequent stages with reoccurrence of tumor and resistance of chemotherapeutic. The recent clinical trails use combinatorial treatment of carboplatin and paclitaxel on ovarian cancer after cytoreduction of tumor. Predominantly patients are responsive initially to therapy and later develop metastases due to drug resistance. Chemotherapy also leads to drug resistance causing enormous variations at cellular level. Multifaceted mechanisms like drug resistance are associated with number of genes and signaling pathways that process the proliferation of cells. Reasons for resistance include epithelial-mesenchyme, DNA repair activation, autophagy, drug efflux, pathway activation, and so on. Determining the routes on molecular mechanism that target chemoresistance pathways are necessary for controlling the treatment and understanding efficient drug targets can open light on improve therapeutic outcomes. Most common drug used for ovarian cancer is Cisplatin, which activates various chemoresistance pathways ultimately causing drug resistance. There have been substantial improvements in understanding the mechanisms of cisplatin resistance or chemo sensitizing cisplatin for effective treatment. Using therapies with combination that involve phytochemical or novel drug delivery system involving the phytochemicals would be a novel treatment in cancer. Phytochemicals are plant-derived compounds that exhibit anticancer, anti-oxidative, anti-inflammatory properties that minimalize side effects exerted from chemotherapeutics.

scholarly journals Predicting Tumor Cell Response to Synergistic Drug Combinations Using a Novel Simplified Deep Learning Model

2020 ◽  
Author(s):  
Heming Zhang ◽  
Jiarui Feng ◽  
Amanda Zeng ◽  
Philip Payne ◽  
Fuhai Li
Keyword(s):  
Deep Learning ◽  
Signaling Pathways ◽  
Drug Combination ◽  
Computational Models ◽  
Learning Model ◽  
Drug Combinations ◽  
Small Set ◽  
Novel Drug ◽  
Deep Learning Model ◽  
Synergistic Drug Combinations

AbstractDrug combinations targeting multiple targets/pathways are believed to be able to reduce drug resistance. Computational models are essential for novel drug combination discovery. In this study, we proposed a new simplified deep learning model, DeepSignalingSynergy, for drug combination prediction. Compared with existing models that use a large number of chemical-structure and genomics features in densely connected layers, we built the model on a small set of cancer signaling pathways, which can mimic the integration of multi-omics data and drug target/mechanism in a more biological meaningful and explainable manner. The evaluation results of the model using the NCI ALMANAC drug combination screening data indicated the feasibility of drug combination prediction using a small set of signaling pathways. Interestingly, the model analysis suggested the importance of heterogeneity of the 46 signaling pathways, which indicates that some new signaling pathways should be targeted to discover novel synergistic drug combinations.

scholarly journals Genomic evaluation of circulating proteins for drug target characterisation and precision medicine

2020 ◽  
Author(s):  
Lasse Folkersen ◽  
Stefan Gustafsson ◽  
Qin Wang ◽  
Daniel Hvidberg Hansen ◽  
Åsa K Hedman ◽  
...  
Keyword(s):  
Precision Medicine ◽  
Drug Target ◽  
Drug Targets ◽  
Clinical Decision Making ◽  
Clinical Decision ◽  
Risk Scores ◽  
Pharmacological Intervention ◽  
Sources Of Information ◽  
C Protein ◽  
Health And Disease

AbstractCirculating proteins are vital in human health and disease and are frequently used as biomarkers for clinical decision-making or as targets for pharmacological intervention. By mapping and replicating protein quantitative trait loci (pQTL) for 90 cardiovascular proteins in over 30,000 individuals, we identified 467 pQTLs for 85 proteins. The pQTLs were used in combination with other sources of information to evaluate known drug targets, and suggest new target candidates or repositioning opportunities, underpinned by a) causality assessment using Mendelian randomization, b) pathway mapping using trans-pQTL gene assignments, and c) protein-centric polygenic risk scores enabling matching of plausible target mechanisms to sub-groups of individuals enabling precision medicine.

Sign in / Sign up

Export Citation Format

Share Document