In Vivo and In Vitro Potency of Polyphosphazene Immunoadjuvants with Hepatitis C Virus Antigen and the Role of Their Supramolecular Assembly.

Two well-defined synthetic polyphosphazene immunoadjuvants, PCPP and PCEP, were studied for their ability to potentiate the immune response to the hepatitis C virus (HCV) E2 glycoprotein antigen in vivo. We report that PCEP induced significantly higher serum neutralization and HCV-specific IgG titers in mice compared to other adjuvants used in the study: PCPP, Alum, and Addavax. PCEP also shifted the response toward the desirable balanced Th1/Th2 immunity, as evaluated by the antibody isotype ratio (IgG2a/IgG1). The in vivo results were analyzed in the context of antigen-adjuvant molecular interactions in the system and in vitro immunostimulatory activity of formulations. Asymmetric flow field flow fractionation (AF4) and dynamic light scattering (DLS) analysis showed that both PCPP and PCEP spontaneously self-assemble with the E2 glycoprotein with the formation of multimeric water-soluble complexes, which demonstrates the role of polyphosphazene macromolecules as vaccine delivery vehicles. Intrinsic in vitro immunostimulatory activity of polyphosphazene adjuvants, which was assessed using a mouse macrophage cell line, revealed comparable activities of both polymers and did not provide an explanation of their in vivo performance. However, PCEP complexes with E2 displayed greater stability against agglomeration and improved in vitro immunostimulatory activity compared to those of PCPP, which is in line with superior in vivo performance of PCEP. The results emphasize the importance of often neglected antigen-polyphosphazene self-assembly mechanisms in formulations, which can provide important insights on their in vivo behavior and facilitate the establishment of a structure-activity relationship for this important class of immunoadjuvants.

Next generation polyphosphazene immunoadjuvant: Synthesis, self-assembly and in vivo potency with human papillomavirus VLPs-based vaccine.

Poly[di(carboxylatomethylphenoxy)phosphazene] (PCMP), a new member of polyphosphazene immunoadjuvant family, is synthesized. In vitro assessment of a new macromolecule revealed hydrolytic degradation profile and immunostimulatory activity comparable to its clinical stage homologue PCPP; however, PCMP was characterized by a beneficial reduced sensitivity to the ionic environment. In vivo evaluation of PCMP potency was conducted with human papillomavirus (HPV) virus-like particles (VLPs) based RG1-VLPs vaccine. In contrast with previously reported self-assembly of polyphosphazene adjuvants with proteins, which typically results in the formation of complexes with multimeric display of antigens, PCMP surface modified VLPs in a composition dependent pattern, which at a high polymer-to VLPs ratio led to stabilization of antigenic particles. Immunization experiments in mice demonstrated that PCMP adjuvanted RG1-VLPs vaccine induced potent humoral immune responses, in particular, on the level of highly desirable protective cross-neutralizing antibodies, and outperformed PCPP and Alhydrogel adjuvanted formulations.

Counting Steps: A New Way to Monitor Patients with Pulmonary Arterial Hypertension.

RATIONALE: Activity levels in patients with pulmonary arterial hypertension (PAH) have correlated with surrogate markers of disease severity. It is not known whether physical activity measures are useful in monitoring patients with PAH. OBJECTIVES: This pilot study aimed to evaluate whether change in physical activity measured by an accelerometer correlates with changes in six-minute walk distance (6MWD), echocardiographic parameters, NT-proBNP, or health-related quality-of-life measures (HRQOL). METHODS: The study design was a prospective, observational study in subjects with prevalent PAH. Subjects wore a wrist-worn accelerometer (Fitbit Charge HR®) between two outpatient visits. Daily step count and activity levels were recorded, and the change over time was correlated with changes in 6MWD, echocardiographic parameters, HRQOL, and NT-proBNP. MEASUREMENTS AND MAIN RESULTS: 30 subjects were enrolled, of which 20 patients had adequate accelerometer data to be analyzed over the study duration. The mean duration of follow-up was 136.4 ( ± 47.3) days. The change in daily step count correlated with a change in 6MWD (r 0.43, p 0.05). Changes in duration spent in moderately active (r 0.52, p 0.02), lightly active (r 0.48, p 0.05), and sedentary activity levels (r - 0.54, p 0.02) correlated with a change in HRQOL. Changes in activity levels did not correlate with echocardiographic measures or NT-pro BNP. CONCLUSIONS: Changes in daily step count and time spent at fairly active, lightly active, and sedentary activity levels correlate with changes in 6MWD, and HRQOL in subjects with PAH suggesting that accelerometry may be a useful monitoring tool.

Macromolecule-Network Electrostatics Controlling Delivery of the Biotherapeutic Cell Modulator TIMP-2.

Tissue inhibitor of metalloproteinase 2 (TIMP-2) is an endogenous 22 kDa proteinase inhibitor, demonstrating antitumorigenic, antimetastatic and antiangiogenic activities in vitro and in vivo. Recombinant TIMP-2 is currently undergoing preclinical testing in multiple, murine tumor models. Here we report the development of an inert, injectable peptide hydrogel matrix enabling encapsulation and sustained release of TIMP-2. We studied the TIMP-2 release profile from four ß-hairpin peptide gels of varying net electrostatic charge. A negatively charged peptide gel (designated AcVES3) enabling encapsulation of 4 mg/mL of TIMP-2, without effects on rheological properties, facilitated the slow sustained release (0.9%/d) of TIMP-2 over 28 d. Released TIMP-2 is structurally intact and maintains the ability to inhibit MMP activity, as well as suppress lung cancer cell proliferation in vitro. These findings suggest that the AcVES3 hydrogel will be useful as an injectable vehicle for systemic delivery of TIMP-2 in vivo for ongoing preclinical development.

Tissue Inhibitor of Metalloprotease-2 (TIMP-2): Bioprocess Development, Physicochemical, Biochemical, and Biological Characterization of Highly Expressed Recombinant Protein.

Tissue inhibitor of metalloprotease-2 (TIMP-2) is a secreted 21 kDa multifunctional protein first described as an endogenous inhibitor of matrix metalloproteinases (MMPs) that prevents breakdown of the extracellular matrix often observed in chronic diseases. TIMP-2 diminishes the level of growth factor-mediated cell proliferation in vitro, as well as neoangiogenesis and tumor growth in vivo independent of its MMP inhibitory activity. These physiological properties make TIMP-2 an excellent candidate for further preclinical development as a biologic therapy of cancer. Here we present a straightforward bioprocessing methodology that yields >35 mg/L recombinant human TIMP-2 6XHis-tagged protein (rhTIMP-2) from suspension cultures of HEK-293-F cells. Enhanced rhTIMP-2-6XHis yields were achieved by optimization of both TIMP-2 cDNA codon sequence and cell culture conditions. Using a two-step chromatographic process, we achieved >95% purity with minimal processing losses. Purified rhTIMP-2-6XHis was free of mouse antigen contamination. Circular dichroism spectroscopy indicated a well-folded rhTIMP-2-6XHis that is highly stable and refractory to pH changes. Two-dimensional heteronuclear single-quantum coherence nuclear magnetic resonance of full length rhTIMP-2-6XHis also indicated a monodisperse, well-folded protein preparation. Purified rhTIMP-2-6XHis inhibited MMP-2 enzymatic activity in a dose-dependent fashion with an IC50 of ∼1.4 nM. Pretreatment of A549 lung cancer and JygMC(A) triple-negative breast cancer cells with rhTIMP-2-6XHis in low-nanomolar amounts inhibited EGF-induced proliferation to basal (unstimulated) levels. This study therefore not only offers a robust bioprocess methodology for rhTIMP-2 production but also characterizes critical physicochemical and biological attributes that are useful for monitoring quality control of the production process.

Modeling, molecular dynamics simulation, and mutation validation for structure of cannabinoid receptor 2 based on known crystal structures of GPCRs.

The cannabinoid receptor 2 (CB2) plays an important role in the immune system. Although a few of GPCRs crystallographic structures have been reported, it is still challenging to obtain functional transmembrane proteins and high resolution X-ray crystal structures, such as for the CB2 receptor. In the present work, we used 10 reported crystal structures of GPCRs which had high sequence identities with CB2 to construct homology-based comparative CB2 models. We applied these 10 models to perform a prescreen by using a training set consisting of 20 CB2 active compounds and 980 compounds randomly selected from the National Cancer Institute (NCI) database. We then utilized the known 170 cannabinoid receptor 1 (CB1) or CB2 selective compounds for further validation. Based on the docking results, we selected one CB2 model (constructed by ß1AR) that was most consistent with the known experimental data, revealing that the defined binding pocket in our CB2 model was well-correlated with the training and testing data studies. Importantly, we identified a potential allosteric binding pocket adjacent to the orthosteric ligand-binding site, which is similar to the reported allosteric pocket for sodium ion Na(+) in the A2AAR and the δ-opioid receptor. Our studies in correlation of our data with others suggested that sodium may reduce the binding affinities of endogenous agonists or its analogs to CB2. We performed a series of docking studies to compare the important residues in the binding pockets of CB2 with CB1, including antagonist, agonist, and our CB2 neutral compound (neutral antagonist) XIE35-1001. Then, we carried out 50 ns molecular dynamics (MD) simulations for the CB2 docked with SR144528 and CP55940, respectively. We found that the conformational changes of CB2 upon antagonist/agonist binding were congruent with recent reports of those for other GPCRs. Based on these results, we further examined one known residue, Val113(3.32), and predicted two new residues, Phe183 in ECL2 and Phe281(7.35), that were important for SR144528 and CP55940 binding to CB2. We then performed site-directed mutation experimental study for these residues and validated the predictions by radiometric binding affinity assay.

Timp2 loss-of-function mutation and TIMP2 treatment in murine model of NSCLC: modulation of immunosuppression and oncogenic signaling.

Mounting evidence suggests that the tissue inhibitor of metalloproteinases-2 (TIMP2) can reduce tumor burden and metastasis. However, the demonstration of such anti-tumor activity and associated mechanisms using in vivo tumor models is lacking. The effects of a Timp2 functional mutation and administration of recombinant TIMP2 were examined in both orthotopic and heterotopic murine models of lung cancer using C57Bl/6 syngeneic Lewis Lung 2-luciferase 2 cells (LL2-luc2) cells. Mice harboring a functional mutation of TIMP2 (mT2) display markedly increased primary lung tumor growth, increased mortality, enriched vasculature, and enhanced infiltration of pro-tumorigenic, immunosuppressive myeloid cells. Treatment with recombinant TIMP2 reduced primary tumor growth in both mutant and wild-type (wt) mice. Comparison of transcriptional profiles of lung tissues from tumor-free, wt versus mT2 mice reveals only minor changes. However, lung tumor-bearing mice of both genotypes demonstrate significant genotype-dependent changes in gene expression following treatment with TIMP. In tumor-bearing wt mice, TIMP2 treatment reduced the expression of upstream oncogenic mediators, whereas treatment of mT2 mice resulted in an immunomodulatory phenotype. A heterotopic subcutaneous model generating metastatic pulmonary tumors demonstrated that daily administration of recombinant TIMP2 significantly downregulates the expression of heat shock proteins, suggesting a reduction of cell-stress responses. In summary, we describe how TIMP2 exerts novel, anti-tumor effects in a murine model of lung cancer and that rTIMP2 treatment supports a normalizing effect on the tumor microenvironment. Our findings show that TIMP2 treatment demonstrates significant potential as an adjuvant in the treatment of NSCLC.

Mistic and TarCF as fusion protein partners for functional expression of the cannabinoid receptor 2 in Escherichia coli.

G protein coupled receptors (GPCRs) are key players in signal recognition and cellular communication making them important therapeutic targets. Large-scale production of these membrane proteins in their native form is crucial for understanding their mechanism of action and target-based drug design. Here we report the overexpression system for a GPCR, the cannabinoid receptor subtype 2 (CB2), in Escherichia coli C43(DE3) facilitated by two fusion partners: Mistic, an integral membrane protein expression enhancer at the N-terminal, and TarCF, a C-terminal fragment of the bacterial chemosensory transducer Tar at the C-terminal of the CB2 open reading frame region. Multiple histidine tags were added on both ends of the fusion protein to facilitate purification. Using individual and combined fusion partners, we found that CB2 fusion protein expression was maximized only when both partners were used. Variable growth and induction conditions were conducted to determine and optimize protein expression. More importantly, this fusion protein Mistic-CB2-TarCF can localize into the E. coli membrane and exhibit functional binding activities with known CB2 ligands including CP55,940, WIN55,212-2 and SR144,528. These results indicate that this novel expression and purification system provides us with a promising strategy for the preparation of biologically active GPCRs, as well as general application for the preparation of membrane-bound proteins using the two new fusion partners described.

ADGAN: Attribute-Driven Generative Adversarial Network for Synthesis and Multiclass Classification of Pulmonary Nodules.

Lung cancer is the leading cause of cancer-related deaths worldwide. According to the American Cancer Society, early diagnosis of pulmonary nodules in computed tomography (CT) scans can improve the five-year survival rate up to 70 % with proper treatment planning. In this article, we propose an attribute-driven Generative Adversarial Network (ADGAN) for synthesis and multiclass classification of Pulmonary Nodules. A self-attention U-Net (SaUN) architecture is proposed to improve the generation mechanism of the network. The generator is designed with two modules, namely, self-attention attribute module (SaAM) and a self-attention spatial module (SaSM). SaAM generates a nodule image based on given attributes whereas SaSM specifies the nodule region of the input image to be altered. A reconstruction loss along with an attention localization loss (AL) is used to produce an attention map prioritizing the nodule regions. To avoid resemblance between a generated image and a real image, we further introduce an adversarial loss containing a regularization term based on KL divergence. The discriminator part of the proposed model is designed to achieve the multiclass nodule classification task. Our proposed approach is validated over two challenging publicly available datasets, namely LIDC-IDRI and LUNGX. Exhaustive experimentation on these two datasets clearly indicate that we have achieved promising classification accuracy as compared to other state-of-the-art methods.

An adaptive registration algorithm for zebrafish larval brain images.

BACKGROUND AND OBJECTIVE: Zebrafish (Danio rerio) in their larval stages have grown increasingly popular as excellent vertebrate models for neurobiological research. Researchers can apply various tools in order to decode the neural structure patterns which can aid the understanding of vertebrate brain development. In order to do so, it is essential to map the gene expression patterns to an anatomical reference precisely. However, high accuracy in sample registration is sometimes difficult to achieve due to laboratory- or protocol-dependent variabilities. METHODS: In this paper, we propose an accurate adaptive registration algorithm for volumetric zebrafish larval image datasets using a synergistic combination of attractive Free-Form-Deformation (FFD) and diffusive Demons algorithms. A coarse registration is achieved first for 3D volumetric data using a 3D affine transformation. A localized registration algorithm in form of a B-splines based FFD is applied next on the coarsely registered volume. Finally, the Demons algorithm is applied on this FFD registered volume for achieving fine registration by making the solution noise resilient. RESULTS: Results Experimental procedures are carried out on a number of 72 hpf (hours post fertilization) 3D confocal zebrafish larval datasets. Comparisons with state-of-the-art methods including some ablation studies clearly demonstrate the effectiveness of the proposed method. CONCLUSIONS: Our adaptive registration algorithm significantly aids Zebrafish imaging analysis over current methods for gene expression anatomical mapping, such as Vibe-Z. We believe the proposed solution would be able to overcome the requirement of high quality images which currently limits the applicability of Zebrafish in neuroimaging research.

A Deep Graph Cut Model For 3D Brain Tumor Segmentation.

Brain tumor segmentation plays a key role in tumor diagnosis and surgical planning. In this paper, we propose a solution to the 3D brain tumor segmentation problem using deep learning and graph cut from the MRI data. In particular, the probability maps of a voxel to belong to the object (tumor) and background classes from the UNet are used to improve the energy function of the graph cut. We derive new expressions for the data term, the region term and the weight factor balancing the data term and the region term for individual voxels in our proposed model. We validate the performance of our model on the publicly available BRATS 2018 dataset. Our segmentation accuracy with a dice similarity score of 0.92 is found to be higher than that of the graph cut and the UNet applied in isolation as well as over a number of state of the art approaches.

Together Recognizing, Localizing and Summarizing Actions in Egocentric Videos.

Analysis of egocentric video has recently drawn attention of researchers in the computer vision as well as multimedia communities. In this paper, we propose a weakly supervised superpixel level joint framework for localization, recognition and summarization of actions in an egocentric video. We first recognize and localize single as well as multiple action(s) in each frame of an egocentric video and then construct a summary of these detected actions. The superpixel level solution helps in precise localization of actions in addition to improving the recognition accuracy. Superpixels are extracted within the central regions of the egocentric video frames; these central regions being determined through a previously developed center-surround model. A sparse spatio-temporal video representation graph is constructed in the deep feature space with the superpixels as nodes. A weakly supervised solution using random walks yields action labels for each superpixel. After determining action label(s) for each frame from its constituent superpixels, we apply a fractional knapsack type formulation for obtaining a summary (of actions). Experimental comparisons on publicly available ADL, GTEA, EGTEA Gaze+, EgoGesture, and EPIC-Kitchens datasets show the effectiveness of the proposed solution.

Nano-Assembly of Quisinostat and Biodegradable Macromolecular Carrier Results in Supramolecular Complexes with Slow-Release Capabilities.

Self-assembly of ionically charged small molecule drugs with water-soluble biodegradable polyelectrolytes into nano-scale complexes can potentially offer a novel and attractive approach to improving drug solubility and prolonging its half-life. Nanoassemblies of quisinostat with water-soluble PEGylated anionic polyphosphazene were prepared by gradient-driven escape of solvent resulting in the reduction of solvent quality for a small molecule drug. A study of binding, analysis of composition, stability, and release profiles was conducted using asymmetric flow field flow fractionation (AF4) and dynamic light scattering (DLS) spectroscopy. Potency assays were performed with WM115 human melanoma and A549 human lung cancer cell lines. The resulting nano-complexes contained up to 100 drug molecules per macromolecular chain and displayed excellent water-solubility and improved hemocompatibility when compared to co-solvent-based drug formulations. Quisinostat release time (complex dissociation) at near physiological conditions in vitro varied from 5 to 14 days depending on initial drug loading. Multimeric complexes displayed dose-dependent potency in cell-based assays and the results were analyzed as a function of complex concentration, as well as total content of drug in the system. The proposed self-assembly process may present a simple alternative to more sophisticated delivery modalities, namely chemically conjugated prodrug systems and nanoencapsulation-based formulations.

Copper-binding anticancer peptides from the piscidin family: an expanded mechanism that encompasses physical and chemical bilayer disruption.

In the search for novel broad-spectrum therapeutics to fight chronic infections, inflammation, and cancer, host defense peptides (HDPs) have garnered increasing interest. Characterizing their biologically-active conformations and minimum motifs for function represents a requisite step to developing them into efficacious and safe therapeutics. Here, we demonstrate that metallating HDPs with Cu2+ is an effective chemical strategy to improve their cytotoxicity on cancer cells. Mechanistically, we find that prepared as Cu2+-complexes, the peptides not only physically but also chemically damage lipid membranes. Our testing ground features piscidins 1 and 3 (P1/3), two amphipathic, histidine-rich, membrane-interacting, and cell-penetrating HDPs that are α-helical bound to membranes. To investigate their membrane location, permeabilization effects, and lipid-oxidation capability, we employ neutron reflectometry, impedance spectroscopy, neutron diffraction, and UV spectroscopy. While P1-apo is more potent than P3-apo, metallation boosts their cytotoxicities by up to two- and seven-fold, respectively. Remarkably, P3-Cu2+ is particularly effective at inserting in bilayers, causing water crevices in the hydrocarbon region and placing Cu2+ near the double bonds of the acyl chains, as needed to oxidize them. This study points at a new paradigm where complexing HDPs with Cu2+ to expand their mechanistic reach could be explored to design more potent peptide-based anticancer therapeutics.

MDL-IWS: Multi-view Deep Learning with Iterative Watershed for Pulmonary Fissure Segmentation.

Pulmonary fissure segmentation is important for localization of lung lesions which include nodules at respective lobar territories. This can be very useful for diagnosis as well as treatment planning. In this paper, we propose a novel coarse-to-fine fissure segmentation approach by proposing a Multi-View Deep Learning driven Iterative WaterShed Algorithm (MDL-IWS). Coarse fissure segmentation obtained from multi-view deep learning yields incomplete fissure volume of interest (VOI) with additional false positives. An iterative watershed algorithm (IWS) is presented to achieve fine segmentation of fissure surfaces. As a part of the IWS algorithm, surface fitting is used to generate a more accurate fissure VOI with substantial reduction in false positives. Additionally, a weight map is used to reduce the over-segmentation of watershed in subsequent iterations. Experiments on the publicly available LOLA11 dataset clearly reveal that our method outperforms several state-of-the-art competitors.

Supramolecular assembly of Toll-like receptor 7/8 agonist into multimeric water-soluble constructs enables superior immune stimulation in vitro and in vivo.

Resiquimod or R848 (RSQD) is a Toll-like receptor (TLR) 7/8 agonist which shows promise as vaccine adjuvant due to its potential to promote highly desirable cellular immunity. The development of this small molecule in the field to date has been largely impeded by its rapid in vivo clearance and lack of association with vaccine antigens. Here, we report a multimeric TLR 7/8 construct of nano-scale size, which results from a spontaneous self-assembly of RSQD with a water-soluble clinical-stage polymer - poly[di(carboxylatophenoxy)phosphazene] (PCPP). The formation of ionically paired construct (PCPP-R) and a ternary complex, which also includes Hepatitis C virus (HCV) antigen, has been demonstrated by dynamic lights scattering (DLS), turbidimetry, fluorescence spectroscopy, asymmetric flow field flow fractionation (AF4), and 1H NMR spectroscopy methods. The resulting supramolecular assembly PCPP-R enabled superior immunostimulation in cellular assays (mouse macrophage reporter cell line) and displayed improved in vitro hemocompatibility (human erythrocytes). In vivo studies demonstrated that PCPP-R adjuvanted HCV formulation induced higher serum neutralization titers in BALB/c mice and shifted the response towards desirable cellular immunity, as evaluated by antibody isotype ratio (IgG2a/IgG1) and ex vivo analysis of cytokine secreting splenocytes (higher levels of interferon gamma (IFN-γ) single and tumor necrosis factor alpha (TNF-α)/IFN-γ double producing cells). The non-covalent multimerization approach stands in contrast to previously suggested RSQD delivery methods, which involve covalent conjugation or encapsulation, and offers a flexible methodology that can be potentially integrated with other parenterally administered drugs.

CHAMPS: Cardiac health Hypergraph Analysis using Multimodal Physiological Signals.

State-of-the-art methods have reported various features for the non-invasive screening of Coronary Artery Disease (CAD). In this paper, we propose a novel approach to represent such features extracted from multiple physiological signals using hypergraph. Firstly, the biological and statistical interconnections among Photoplethysmogram (PPG) and Phonocardiogram (PCG) features are exploited by connecting them as hyperedges. Then, metadata features (age, weight and height) are connected using hyperedges with the rest of the features. Hypergraph based formalism provides greater flexibility in capturing the interrelationships among different features as compared to the graph counterpart. Finally, hypergraph laplacian as a derived feature is applied to classify CAD against non-CAD. The proposed method is validated on PPG and PCG data collected in a hospital setup. The results reveal 98% Sensitivity and 82% Specificity, leading to 92% classification accuracy.

An Iterative Spanning Forest Framework for Superpixel Segmentation.

Superpixel segmentation has emerged as an important research problem in the areas of image processing and computer vision. In this paper, we propose a framework, namely Iterative Spanning Forest (ISF), in which improved sets of connected superpixels (supervoxels in 3D) can be generated by a sequence of image foresting transforms. In this framework, one can choose the most suitable combination of ISF components for a given application-i.e., 1) a seed sampling strategy; 2) a connectivity function; 3) an adjacency relation; and 4) a seed pixel recomputation procedure. The superpixels in ISF structurally correspond to spanning trees rooted at those seeds. We present five ISF-based methods to illustrate different choices for those components. These methods are compared with a number of state-of-the-art approaches with respect to effectiveness and efficiency. Experiments are carried out on several datasets containing 2D and 3D objects with distinct texture and shape properties, including a high-level application, named sky image segmentation. The theoretical properties of ISF are demonstrated in the supplementary material and the results show ISF-based methods rank consistently among the best for all datasets.

CHANGE: Cardiac Health Analysis Using Graph Eigenvalues.

Coronary Artery Disease (CAD) is an important problem in cardiac health and is a leading cause of human mortality. Prior arts have shown that features extracted from non-invasive Photoplethysmogram (PPG) signal are effective in classifying CAD. In this paper, we represent cardiac health as a graph (CHG) in order to exploit the dependencies of PPG features as well as the metadata features. We then compute spectral features from the eigenvalues of the graph Laplacian of CHG. Finally, k-means algorithm is employed for classifying the data into CAD and non-CAD. Unsupervised experiments on a cohort with 32 participants yields 88% accuracy and demonstrates advantage of the proposed formulation over a baseline and two state-of-the-art approaches.