Artificial intelligence in cancer target identification and drug discovery.

Artificial intelligence is an advanced method to identify novel anticancer targets and discover novel drugs from biology networks because the networks can effectively preserve and quantify the interaction between components of cell systems underlying human diseases such as cancer. Here, we review and discuss how to employ artificial intelligence approaches to identify novel anticancer targets and discover drugs. First, we describe the scope of artificial intelligence biology analysis for novel anticancer target investigations. Second, we review and discuss the basic principles and theory of commonly used network-based and machine learning-based artificial intelligence algorithms. Finally, we showcase the applications of artificial intelligence approaches in cancer target identification and drug discovery. Taken together, the artificial intelligence models have provided us with a quantitative framework to study the relationship between network characteristics and cancer, thereby leading to the identification of potential anticancer targets and the discovery of novel drug candidates.

Graph convolutional network approach to discovering disease-related circRNA-miRNA-mRNA axes.

Non-coding RNAs are gaining prominence in biology and medicine, as they play major roles in cellular homeostasis among which the circRNA-miRNA-mRNA axes are involved in a series of disease-related pathways, such as apoptosis, cell invasion and metastasis. Recently, many computational methods have been developed for the prediction of the relationship between ncRNAs and diseases, which can alleviate the time-consuming and labor-intensive exploration involved with biological experiments. However, these methods handle ncRNAs separately, ignoring the impact of the interactions among ncRNAs on the diseases. In this paper we present a novel approach to discovering disease-related circRNA-miRNA-mRNA axes from the disease-RNA information network. Our method, using graph convolutional network, learns the characteristic representation of each biological entity by propagating and aggregating local neighbor information based on the global structure of the network. The approach is evaluated using the real-world datasets and the results show that it outperforms other state-of-the-art baselines on most of the metrics.

Developing the novel bioinformatics algorithms to systematically investigate the connections among survival time, key genes and proteins for Glioblastoma multiforme.

BACKGROUND: Glioblastoma multiforme (GBM) is one of the most common malignant brain tumors and its average survival time is less than 1 year after diagnosis. RESULTS: Firstly, this study aims to develop the novel survival analysis algorithms to explore the key genes and proteins related to GBM. Then, we explore the significant correlation between AEBP1 upregulation and increased EGFR expression in primary glioma, and employ a glioma cell line LN229 to identify relevant proteins and molecular pathways through protein network analysis. Finally, we identify that AEBP1 exerts its tumor-promoting effects by mainly activating mTOR pathway in Glioma. CONCLUSIONS: We summarize the whole process of the experiment and discuss how to expand our experiment in the future.

Combining deep residual neural network features with supervised machine learning algorithms to classify diverse food image datasets.

Obesity is increasing worldwide and can cause many chronic conditions such as type-2 diabetes, heart disease, sleep apnea, and some cancers. Monitoring dietary intake through food logging is a key method to maintain a healthy lifestyle to prevent and manage obesity. Computer vision methods have been applied to food logging to automate image classification for monitoring dietary intake. In this work we applied pretrained ResNet-152 and GoogleNet convolutional neural networks (CNNs), initially trained using ImageNet Large Scale Visual Recognition Challenge (ILSVRC) dataset with MatConvNet package, to extract features from food image datasets; Food 5K, Food-11, RawFooT-DB, and Food-101. Deep features were extracted from CNNs and used to train machine learning classifiers including artificial neural network (ANN), support vector machine (SVM), Random Forest, and Naive Bayes. Results show that using ResNet-152 deep features with SVM with RBF kernel can accurately detect food items with 99.4% accuracy using Food-5K validation food image dataset and 98.8% with Food-5K evaluation dataset using ANN, SVM-RBF, and Random Forest classifiers. Trained with ResNet-152 features, ANN can achieve 91.34%, 99.28% when applied to Food-11 and RawFooT-DB food image datasets respectively and SVM with RBF kernel can achieve 64.98% with Food-101 image dataset. From this research it is clear that using deep CNN features can be used efficiently for diverse food item image classification. The work presented in this research shows that pretrained ResNet-152 features provide sufficient generalisation power when applied to a range of food image classification tasks.

Analysis of Organization of the Interactome Using Dominating Sets: A Case Study on Cell Cycle Interaction Networks.

In this study, a minimum dominating set based approach was developed and implemented as a Cytoscape plugin to identify critical and redundant proteins in a protein interaction network. We focused on the investigation of the properties associated with critical proteins in the context of the analysis of interaction networks specific to cell cycle in both yeast and human. A total of 132 yeast genes and 129 human proteins have been identified as critical nodes while 950 in yeast and 980 in human have been categorized as redundant nodes. A clear distinction between critical and redundant proteins was observed when examining their topological parameters including betweenness centrality, suggesting a central role of critical proteins in the control of a network. The significant differences in terms of gene coexpression and functional similarity were observed between the two sets of proteins in yeast. Critical proteins were found to be enriched with essential genes in both networks and have a more deleterious effect on the network integrity than their redundant counterparts. Furthermore, we obtained statistically significant enrichments of proteins that govern human diseases including cancer-related and virus-targeted genes in the corresponding set of critical proteins.

Integrating Omic Data with a Multiplex Network-based Approach for the Identification of Cancer Subtypes.

Comprehensive characterization and identification of cancer subtypes have a number of applications and implications in life science and cancer research. Technologies centered on the integration of omics data hold great promise in this endeavor. This paper proposed a multiplex network-based approach for integrative analysis of heterogeneous omics data. It represents a useful alternative network-based solution to the problem and a significant step forward to the methods in which each type of data is treated independently. It has been tested on the identification of the subtypes of glioblastoma multiforme and breast invasive carcinoma from three omics data. The results obtained have shown that it has achieved the performance comparable to state-of-the-art techniques (Normalized Mutual Information > 0.8). In comparison to traditional systems biology tools, the proposed methodology has several significant advantages. It has the ability to correlate and integrate multiple data levels in a holistic manner which may be useful to facilitate our understanding of the pathogenesis of diseases and to capture the heterogeneity of biological processes and the complexity of phenotypes.

Pulmonary epithelial CCR3 promotes LPS-induced lung inflammation by mediating release of IL-8.

Interleukin (IL)-8 from pulmonary epithelial cells has been suggested to play an important role in the airway inflammation, although the mechanism remains unclear. We envisioned a possibility that pulmonary epithelial CCR3 could be involved in secretion and regulation of IL-8 and promote lipopolysaccharide (LPS)-induced lung inflammation. Human bronchial epithelial cell line NCI-H292 and alveolar type II epithelial cell line A549 were used to test role of CCR3 in production of IL-8 at cellular level. In vivo studies were performed on C57/BL6 mice instilled intratracheally with LPS in a model of acute lung injury (ALI). The activity of a CCR3-specific inhibitor (SB-328437) was measured in both in vitro and in vivo systems. We found that expression of CCR3 in NCI-H292 and A549 cells were increased by 23% and 16%, respectively, 24 h after the challenge with LPS. LPS increased the expression of CCR3 in NCI-H292 and A549 cells in a time-dependent manner, which was inhibited significantly by SB-328437. SB-328437 also diminished neutrophil recruitment in alveolar airspaces and improved LPS-induced ALI and production of IL-8 in bronchoalveolar lavage fluid. These results suggest that pulmonary epithelial CCR3 be involved in progression of LPS-induced lung inflammation by mediating release of IL-8. CCR3 in pulmonary epithelia may be an attractive target for development of therapies for ALI.

Predictive integration of gene functional similarity and co-expression defines treatment response of endothelial progenitor cells.

BACKGROUND: Endothelial progenitor cells (EPCs) have been implicated in different processes crucial to vasculature repair, which may offer the basis for new therapeutic strategies in cardiovascular disease. Despite advances facilitated by functional genomics, there is a lack of systems-level understanding of treatment response mechanisms of EPCs. In this research we aimed to characterize the EPCs response to adenosine (Ado), a cardioprotective factor, based on the systems-level integration of gene expression data and prior functional knowledge. Specifically, we set out to identify novel biosignatures of Ado-treatment response in EPCs. RESULTS: The predictive integration of gene expression data and standardized functional similarity information enabled us to identify new treatment response biosignatures. Gene expression data originated from Ado-treated and -untreated EPCs samples, and functional similarity was estimated with Gene Ontology (GO)-based similarity information. These information sources enabled us to implement and evaluate an integrated prediction approach based on the concept of k-nearest neighbours learning (kNN). The method can be executed by expert- and data-driven input queries to guide the search for biologically meaningful biosignatures. The resulting integrated kNN system identified new candidate EPC biosignatures that can offer high classification performance (areas under the operating characteristic curve>0.8). We also showed that the proposed models can outperform those discovered by standard gene expression analysis. Furthermore, we report an initial independent in vitro experimental follow-up, which provides additional evidence of the potential validity of the top biosignature. CONCLUSION: Response to Ado treatment in EPCs can be accurately characterized with a new method based on the combination of gene co-expression data and GO-based similarity information. It also exploits the incorporation of human expert-driven queries as a strategy to guide the automated search for candidate biosignatures. The proposed biosignature improves the systems-level characterization of EPCs. The new integrative predictive modeling approach can also be applied to other phenotype characterization or biomarker discovery problems.

Systems-based biological concordance and predictive reproducibility of gene set discovery methods in cardiovascular disease.

The discovery of novel disease biomarkers is a crucial challenge for translational bioinformatics. Demonstration of both their classification power and reproducibility across independent datasets are essential requirements to assess their potential clinical relevance. Small datasets and multiplicity of putative biomarker sets may explain lack of predictive reproducibility. Studies based on pathway-driven discovery approaches have suggested that, despite such discrepancies, the resulting putative biomarkers tend to be implicated in common biological processes. Investigations of this problem have been mainly focused on datasets derived from cancer research. We investigated the predictive and functional concordance of five methods for discovering putative biomarkers in four independently-generated datasets from the cardiovascular disease domain. A diversity of biosignatures was identified by the different methods. However, we found strong biological process concordance between them, especially in the case of methods based on gene set analysis. With a few exceptions, we observed lack of classification reproducibility using independent datasets. Partial overlaps between our putative sets of biomarkers and the primary studies exist. Despite the observed limitations, pathway-driven or gene set analysis can predict potentially novel biomarkers and can jointly point to biomedically-relevant underlying molecular mechanisms.

Differential impairment of regulatory T cells rather than effector T cells by paclitaxel-based chemotherapy.

Characterized as a mitotic inhibitor, paclitaxel has gained importance as a promising agent for the treatment of advanced non-small cell lung cancer (NSCLC). However, whether paclitaxel has immune modulatory effects remains unclear. In this study, we analyzed 55 peripheral blood samples from NSCLC patients who underwent paclitaxel-based chemotherapy. We found that among the lymphocyte subsets, paclitaxel selectively decreased the size of the regulatory T cell (Treg) population rather than other subsets including effector T cells (Teff). Apoptosis by upregulating the expression of the cell death receptor Fas (CD95) contributed to the reduced cell number of Treg. Importantly, the inhibitory function of Treg was significantly impaired, while the production of Th1 cytokines IFN-gamma and IL-2 and the expression of the activation marker CD44 among CD4(+) and CD8(+) T cells were augmented after paclitaxel treatment. These results strongly demonstrated that paclitaxel-based chemotherapy played important roles in modulating immune responses.

Clustering-based approaches to SAGE data mining.

Serial analysis of gene expression (SAGE) is one of the most powerful tools for global gene expression profiling. It has led to several biological discoveries and biomedical applications, such as the prediction of new gene functions and the identification of biomarkers in human cancer research. Clustering techniques have become fundamental approaches in these applications. This paper reviews relevant clustering techniques specifically designed for this type of data. It places an emphasis on current limitations and opportunities in this area for supporting biologically-meaningful data mining and visualisation.

A computational approach for peptidomic analysis in taxonomic study of amphibian species.

Peptides in the skin secretion of frogs have been studied for some time now because they frequently possess important biological activity such as antibiotic, antimicrobial, or anticancer properties. In this paper, we present a computational approach for measuring the degree of similarity between the entire peptide complement of the skin secretion of specimens from the same or different species. The first step in the analysis is the generation of a mass spectral profile from an experimental high-performance liquid chromatography/electrosparay ionization analysis of the sample. An "overlap" between the mass spectral profiles of different specimens is then proposed as a measure of their similarity. Analysis of specimens from three species of the genus Litoria, viz., L. Aurea, L. Caerulea, and L. Infrafrenata, and Rana Capito of genus Rana shows that the degree of similarity is highest between specimens from the same species, lower for specimens from different species of the same genus, and lowest between specimens from different genera. This indicates that comparison of skin peptide profiles (i.e., mass spectral profiles of skin secretion) is potentially a useful aid in the taxonomic study of amphibian species.

Cross-sectional area reduction of the aortic ostium by suprarenal stent wires: in vitro phantom study by CT virtual angioscopy.

The study aims to investigate the reduction of cross-sectional area of the aortic ostium by the presence of aortic stent wires observed using CT virtual angioscopy in an aorta phantom. A human aorta phantom was built with a commercial stent graft placed in situ to simulate a repaired aortic aneurysm. Virtual angioscopic images of the aortic ostium and stent wires were generated in the locations of renal arteries, superior mesenteric artery and corresponding cross-sectional area reduction caused by stent wires was measured by virtual angioscopy in various scanning parameters. Our study showed that cross-sectional area reduction of the aortic ostium was determined by the diameter of renal ostium and stent wires, as well as the number of stent wires crossing the aortic ostium.

Heuristic charge assignment for deconvolution of electrospray ionization mass spectra.

We propose a new algorithm for deconvolution of electrospray ionization mass spectra based on direct assignment of charge to the measured signal at each mass-to-charge ratio (m/z). We investigate two heuristics for charge assignment: the entropy-based heuristic is adapted from a deconvolution algorithm by Reinhold and Reinhold;10 the multiplicative-correlation heuristic is adapted from the multiplicative-correlation deconvolution algorithm of Hagen and Monnig.6 The entropy-based heuristic is insensitive to overestimates of z(max), the maximum ion charge. We test the deconvolution algorithm on two single-component samples: the measured spectrum of human beta-endorphin has two prominent and one very weak line whereas myoglobin has a well-developed quasi-gaussian envelope of 17 peaks. In both cases, the deconvolution algorithm gives a clean deconvoluted spectrum with one dominant peak and very few artefacts. The relative heights of the peaks due to the parent molecules in the deconvoluted spectrum of a mixture of two peptides, which are expected to ionize with equal efficiency, give an accurate measure of their relative concentration in the sample.