Dimensional Descriptors Research Articles

Mapper–Type Algorithms for Complex Data and Relations

Mapper and Ball Mapper are Topological Data Analysis tools used for exploring high dimensional point clouds and visualizing scalar–valued functions on those point clouds. Inspired by open questions in knot theory, new features are added to Ball Mapper that enable encoding of the structure, internal relations and symmetries of the point cloud. Moreover, the strengths of Mapper and Ball Mapper constructions are combined to create a tool for comparing high dimensional data descriptors of a single dataset. This new hybrid algorithm, Mapper on Ball Mapper, is applicable to high dimensional lens functions. As a proof of concept we include applications to knot and game theory.

Informative Training Data for Efficient Property Prediction in Metal-Organic Frameworks by Active Learning.

In recent data-driven approaches to material discovery, scenarios where target quantities are expensive to compute and measure are often overlooked. In such cases, it becomes imperative to construct a training set that includes the most diverse, representative, and informative samples. Here, a novel regression tree-based active learning algorithm is employed for such a purpose. It is applied to predict the band gap and adsorption properties of metal-organic frameworks (MOFs), a novel class of materials that results from the virtually infinite combinations of their building units. Simpler and low dimensional descriptors, such as those based on stoichiometric and geometric properties, are used to compute the feature space for this model owing to their ability to better represent MOFs in the low data regime. The partitions given by a regression tree constructed on the labeled part of the data set are used to select new samples to be added to the training set, thereby limiting its size while maximizing the prediction quality. Tests on the QMOF, hMOF, and dMOF data sets reveal that our method constructs small training data sets to learn regression models that predict the target properties more efficiently than existing active learning approaches, and with lower variance. Specifically, our active learning approach is highly beneficial when labels are unevenly distributed in the descriptor space and when the label distribution is imbalanced, which is often the case for real world data. The regions defined by the tree help in revealing patterns in the data, thereby offering a unique tool to efficiently analyze complex structure-property relationships in materials and accelerate materials discovery.

Positional difference and Frequency (PdF) based alignment-free technique for genome sequence comparison

In the field of computational biology, genome sequence comparison among different species is essential and has applications in both the research and scientific fields. Owing to the lengthy processing time and large number of data sets, the alignment-based approaches are unsuitable and ineffective. Therefore, alignment-free techniques have obtained popularity for acquiring proper sequence clustering and evolutionary relationship among species. In this paper, a complete bipartite graph based Positional difference and Frequency (PdF) vector descriptor is introduced. Positional difference and Frequency, two parameters, are applied to the genome sequence to create a 16- dimensional vector descriptor using the di-nucleotide representation of genome sequence. Subsequently, a distance matrix is calculated to construct the phylogenetic trees for different data sets of mammals and viruses. The achieved outcomes are compared with the phylogenetic trees of the earlier methods viz. the FFP method, the ClustalW method, the MEV method, the PCNV method and the FIS method. In most instances, the proposed method produces more precise outcomes than the preceding techniques and has potential for genome sequence comparison on both the equal and unequal length of data-sets. Communicated by Ramaswamy H. Sarma

The non-ergodic nature of mental health and psychiatric disorders: implications for biomarker and diagnostic research.

The non-ergodic nature of mental health and psychiatric disorders: implications for biomarker and diagnostic research.

Can molecular dynamics simulations improve predictions of protein-ligand binding affinity with machine learning?

Binding affinity prediction largely determines the discovery efficiency of lead compounds in drug discovery. Recently, machine learning (ML)-based approaches have attracted much attention in hopes of enhancing the predictive performance of traditional physics-based approaches. In this study, we evaluated the impact of structural dynamic information on the binding affinity prediction by comparing the models trained on different dimensional descriptors, using three targets (i.e. JAK1, TAF1-BD2 and DDR1) and their corresponding ligands as the examples. Here, 2D descriptors are traditional ECFP4 fingerprints, 3D descriptors are the energy terms of the Smina and NNscore scoring functions and 4D descriptors contain the structural dynamic information derived from the trajectories based on molecular dynamics (MD) simulations. We systematically investigate the MD-refined binding affinity prediction performance of three classical ML algorithms (i.e. RF, SVR and XGB) as well as two common virtual screening methods, namely Glide docking and MM/PBSA. The outcomes of the ML models built using various dimensional descriptors and their combinations reveal that the MD refinement with the optimized protocol can improve the predictive performance on the TAF1-BD2 target with considerable structural flexibility, but not for the less flexible JAK1 and DDR1 targets, when taking docking poses as the initial structure instead of the crystal structures. The results highlight the importance of the initial structures to the final performance of the model through conformational analysis on the three targets with different flexibility.

Efficient Interrogation of the Kinetic Barriers Demarcating Catalytic States of a Tyrosine Kinase with Optimal Physical Descriptors and Mixture Models.

Computer simulations are increasingly used to access thermo-kinetic information underlying structural transformation of protein kinases. Such information are necessary to probe their roles in disease progression and interactions with drug targets. However, the investigations are frequently challenged by forbiddingly high computational expense, and by the lack of standard protocols for the design of low dimensional physical descriptors that encode system features important for transitions. Here, we consider the demarcating characteristics of the different states of Abelson tyrosine kinase associated with distinct catalytic activity to construct a set of physically meaningful, orthogonal collective variables that preserve the slow modes of the system. Independent sampling of each metastable state is followed by the estimation of global partition function along the appropriate physical descriptors using the modified Expectation Maximized Molecular Dynamics method. The resultant free energy barriers are in excellent agreement with experimentally known rate-limiting dynamics and activation energy computed with conventional enhanced sampling methods. We discuss possible directions for further development and applications.

An unsupervised framework for extracting multilane roads from OpenStreetMap

Multilane roads are a set of approximately parallel line segments representing the same road in large-scale vector maps. They must be extracted first in cartographic generalization. There are numerous multilane roads in the easily accessible OpenStreetMap (OSM) dataset. For this dataset, polygon-based methods have achieved state-of-the-art performance. However, traditional polygon-based methods usually rely on manually labeled data, which means they are time-consuming and labor-intensive. To address this problem, an unsupervised framework for extracting multilane roads is proposed in this study. Road segments were first grouped to form the road polygons. A set of shape descriptors was formulated to reduce the dimensions of individual road polygons into conceptual points. Next, dimensional shape descriptors were standardized using logarithmic standardization. The density peaks clustering (DPC) algorithm was employed to classify these points. Then, cluster tags were identified manually to recognize which clusters represent multilane polygons. Finally, post-processing learning from the concept of assimilation is proposed to fill holes and remove islands. Experiments were conducted to extract multilane roads with datasets from three cities: Wuhan, Beijing and Munich. The experimental results show that the proposed framework effectively extracted multilane roads without any labels with accuracy levels comparable to those of supervised methods.

Two Dimensional Descriptors Based on Degree, Neighborhood Degree, and Reverse Degree for HEX (Hexagonal) Lattice

Crystal structures are of great scrutiny due to the elegant and well‐ordered symmetry that influences a significant role in determining numerous physical properties. Our aim is to perceive the role of topological descriptors in the field of crystallography using chemical graph theory to examine symmetrical crystal structure HEX. Simple hexagonal (HEX) is a crystal structure formed by arranging the same layer of atoms in a hexagon with one additional atom at the center. Chemical graph theory allows us to study a variety of molecular structures via graphical representation where each atom is denoted as a vertex and the bond form between them is defined as edge. In this research work, we compute the general Randi index, atom bond connectivity index, geometric arithmetic index, first and second Zagreb indices. Furthermore, we will compute their neighborhood and reverse degree‐based versions and visualize which descriptor stands high in accordance with its numerical value.

Jointly Aligning and Predicting Continuous Emotion Annotations.

Time-continuous dimensional descriptions of emotions (e.g., arousal, valence) allow researchers to characterize short-time changes and to capture long-term trends in emotion expression. However, continuous emotion labels are generally not synchronized with the input speech signal due to delays caused by reaction-time, which is inherent in human evaluations. To deal with this challenge, we introduce a new convolutional neural network (multi-delay sinc network) that is able to simultaneously align and predict labels in an end-to-end manner. The proposed network is a stack of convolutional layers followed by an aligner network that aligns the speech signal and emotion labels. This network is implemented using a new convolutional layer that we introduce, the delayed sinc layer. It is a time-shifted low-pass (sinc) filter that uses a gradient-based algorithm to learn a single delay. Multiple delayed sinc layers can be used to compensate for a non-stationary delay that is a function of the acoustic space. We test the efficacy of this system on two common emotion datasets, RECOLA and SEWA, and show that this approach obtains state-of-the-art speech-only results by learning time-varying delays while predicting dimensional descriptors of emotions.

Molecular descriptor analysis of approved drugs using unsupervised learning for drug repurposing

Machine learning and data-driven approaches are currently being widely used in drug discovery and development due to their potential advantages in decision-making based on the data leveraged from existing sources. Applying these approaches to drug repurposing (DR) studies can identify new relationships between drug molecules, therapeutic targets and diseases that will eventually help in generating new insights for developing novel therapeutics. In the current study, a dataset of 1671 approved drugs is analyzed using a combined approach involving unsupervised Machine Learning (ML) techniques (Principal Component Analysis (PCA) followed by k-means clustering) and Structure-Activity Relationships (SAR) predictions for DR. PCA is applied on all the two dimensional (2D) molecular descriptors of the dataset and the first five Principal Components (PC) were subsequently used to cluster the drugs into nine well separated clusters using k-means algorithm. We further predicted the biological activities for the drug-dataset using the PASS (Predicted Activities Spectra of Substances) tool. These predicted activity values are analyzed systematically to identify repurposable drugs for various diseases. Clustering patterns obtained from k-means showed that every cluster contains subgroups of structurally similar drugs that may or may not have similar therapeutic indications. We hypothesized that such structurally similar but therapeutically different drugs can be repurposed for the native indications of other drugs of the same cluster based on their high predicted biological activities obtained from PASS analysis. In line with this, we identified 66 drugs from the nine clusters which are structurally similar but have different therapeutic uses and can therefore be repurposed for one or more native indications of other drugs of the same cluster. Some of these drugs not only share a common substructure but also bind to the same target and may have a similar mechanism of action, further supporting our hypothesis. Furthermore, based on the analysis of predicted biological activities, we identified 1423 drugs that can be repurposed for 366 new indications against several diseases. In this study, an integrated approach of unsupervised ML and SAR analysis have been used to identify new indications for approved drugs and the study provides novel insights into clustering patterns generated through descriptor level analysis of approved drugs.

Approach for 3D Cultural Relic Classification Based on a Low-Dimensional Descriptor and Unsupervised Learning

Computer-aided classification serves as the basis of virtual cultural relic management and display. The majority of the existing cultural relic classification methods require labelling of the samples of the dataset; however, in practical applications, there is often a lack of category labels of samples or an uneven distribution of samples of different categories. To solve this problem, we propose a 3D cultural relic classification method based on a low dimensional descriptor and unsupervised learning. First, the scale-invariant heat kernel signature (Si-HKS) was computed. The heat kernel signature denotes the heat flow of any two vertices across a 3D shape and the heat diffusion propagation is governed by the heat equation. Secondly, the Bag-of-Words (BoW) mechanism was utilized to transform the Si-HKS descriptor into a low-dimensional feature tensor, named a SiHKS-BoW descriptor that is related to entropy. Finally, we applied an unsupervised learning algorithm, called MKDSIF-FCM, to conduct the classification task. A dataset consisting of 3D models from 41 Tang tri-color Hu terracotta Eures was utilized to validate the effectiveness of the proposed method. A series of experiments demonstrated that the SiHKS-BoW descriptor along with the MKDSIF-FCM algorithm showed the best classification accuracy, up to 99.41%, which is a solution for an actual case with the absence of category labels and an uneven distribution of different categories of data. The present work promotes the application of virtual reality in digital projects and enriches the content of digital archaeology.

Human Body Shape Reconstruction from Binary Image Using Convolutional Neural Network

Human body modeling is an important part of virtual try-on. In order to quickly reconstruct the three-dimensional(3D) human body based on the minimum input, we propose a new method for accurate reconstruction of the human body by inputting binary images from either single view or multiple views. We first encode the shape of the human body via Principal Component Analysis (PCA) to extract the low dimensional shape descriptor. Secondly, we design a novel Body Reconstruction Convolutional Neural Network (BRCNN) with two branches, which could capture deep correlated features from different views and merge them. Given the obtained statistical shape space of the human body, we jointly train the BRCNN to learn a global mapping from the input to the shape descriptor which can be then decoded to points cloud for the reconstruction of various body shapes under neutral poses. The experimental results show that compared with the existing human reconstruction technology, the accuracy has been improved by 1.07%, and the prediction results of the two views are better than those from the single view. Further investigation also reveals that the prediction results of the weight-sharing network are better than the network without weight-sharing.

Large-scale evaluation of cytochrome P450 2C9 mediated drug interaction potential with machine learning-based consensus modeling

Cytochrome P450 (CYP) enzymes play an important role in the metabolism of xenobiotics. Since they are connected to drug interactions, screening for potential inhibitors is of utmost importance in drug discovery settings. Our study provides an extensive classification model for P450-drug interactions with one of the most prominent members, the 2C9 isoenzyme. Our model involved the largest set of 45,000 molecules ever used for developing prediction models. The models are based on three different types of descriptors, (a) typical one, two and three dimensional molecular descriptors, (b) chemical and pharmacophore fingerprints and (c) interaction fingerprints with docking scores. Two machine learning algorithms, the boosted tree and the multilayer feedforward of resilient backpropagation network were used and compared based on their performances. The models were validated both internally and using external validation sets. The results showed that the consensus voting technique with custom probability thresholds could provide promising results even in large-scale cases without any restrictions on the applicability domain. Our best model was capable to predict the 2C9 inhibitory activity with the area under the receiver operating characteristic curve (AUC) of 0.85 and 0.84 for the internal and the external test sets, respectively. The chemical space covered with the largest available dataset has reached its limit encompassing publicly available bioactivity data for the 2C9 isoenzyme.

Artificial Intelligence for Cardiac Imaging-Genetics Research.

Cardiovascular conditions remain the leading cause of mortality and morbidity worldwide, with genotype being a significant influence on disease risk. Cardiac imaging-genetics aims to identify and characterize the genetic variants that influence functional, physiological, and anatomical phenotypes derived from cardiovascular imaging. High-throughput DNA sequencing and genotyping have greatly accelerated genetic discovery, making variant interpretation one of the key challenges in contemporary clinical genetics. Heterogeneous, low-fidelity phenotyping and difficulties integrating and then analyzing large-scale genetic, imaging and clinical datasets using traditional statistical approaches have impeded process. Artificial intelligence (AI) methods, such as deep learning, are particularly suited to tackle the challenges of scalability and high dimensionality of data and show promise in the field of cardiac imaging-genetics. Here we review the current state of AI as applied to imaging-genetics research and discuss outstanding methodological challenges, as the field moves from pilot studies to mainstream applications, from one dimensional global descriptors to high-resolution models of whole-organ shape and function, from univariate to multivariate analysis and from candidate gene to genome-wide approaches. Finally, we consider the future directions and prospects of AI imaging-genetics for ultimately helping understand the genetic and environmental underpinnings of cardiovascular health and disease.

Three-dimensional foot shape analysis in children: a pilot analysis using three-dimensional shape descriptors

BackgroundExisting clinical measures to describe foot morphology are limited in that they are commonly two-dimensional, low in resolution and accuracy, and do not accurately represent the multi-planar and complex changes during development across childhood. Using three-dimensional (3D) scanner technology provides the opportunity to understand more about morphological changes throughout childhood with higher resolution and potentially more relevant 3D shape measures. This is important to advance the prevailing arguments about the typical development of children’s feet and inform the development of appropriate clinical measures. 3D shape descriptors derived from 3D scanning can be used to quantify changes in shape at each point of the 3D surface. The aim of this study was to determine whether 3D shape descriptors derived from 3D scanning data can identify differences in foot morphology between children of different ages.MethodsFifteen children were recruited from three age groups (2, 5, and 7 years of age). Both feet were scanned in bipedal stance, using the Artec Eva (Artec Group, Luxembourg, Luxembourg) hand-held scanner. Three dimensional shape descriptors were extracted from the 3D scans of the right foot, to create histograms for each age group and heat maps of representative participants for comparison.ResultsThere were changes to the dorsal, medial and lateral surfaces of the feet with age. The surfaces became less round along with an increase in indented areas. This is supported by the heat maps which demonstrated that the surfaces of the anatomical landmarks (e.g. the malleoli and navicular tuberosity) became more rounded and protruding, with indented surfaces appearing around these landmarks. On the plantar surface, the concavity of the midfoot was evident and this concavity extended into the midfoot from the medial aspect as age increased.ConclusionsThe findings of this study indicated that with increasing age the foot becomes thinner in 3D, with bony architecture emerging, and the medial longitudinal arch (MLA) increases in area and concavity. Three-dimensional shape descriptors have shown good potential for locating and quantifying changes in foot structure across childhood. Three-dimensional shape descriptor data will be beneficial for understanding more about foot development and quantifying changes over time.

Morphometric Analysis of Biasterion Length and Cranial Vault Weight of Nigerian Skulls

Anthropometry involves the systematic measurement of the physical properties of the human body, primarily dimensional descriptors of body size and shape. It plays an important role in industrial design, clothing, ergonomics and architecture. Skull is important in this regard as it resists adverse environmental conditions over time. The present study aims to determine the anthropometric dimensions of biasterion length and cranial vault weight of Nigerian skulls. One hundred adult dry skulls, (82 males, and 18 females) free from damage and deformities collected from eleven Departments of Anatomy in Nigerian Universities were used. Tape and weighing scale were used to measure biasterion length and weight of cranial vault. Data was analyzed with Graph Pad Prism 5.03. The mean±SEM, maximum, minimum and sexual dimorphism ratio were computed. The mean of biasterion length (mm) of male and female were 127.0±0.90 and 123.0±1.56 mm respectively while the weight of cranial vault for male and female were 330±6.08 and 308±7.84 respectively. The result of this study will be of importance in providing new perspectives on human osteology and in sex determination in forensic anthropology.

LPD-AE: Latent Space Representation of Large-Scale 3D Point Cloud

The effective latent space representation of point cloud provides a foremost and fundamental manner that can be used for challenging tasks, including point cloud based place recognition and reconstruction, especially in large-scale dynamic environments. In this paper, we present a novel deep neural network, LPD-AE(Large-scale Place Description AutoEncoder Network), to obtain meaningful local and contextual features for the generation of latent space from 3D point cloud directly. The encoder network constructs the discriminative global descriptors to realize high accuracy and robust place recognition, which contributed by extracting the local neighbor geometric features and aggregating neighborhood relationships both in feature space and physical space. The decoder network performs hierarchical reconstruction on coarse key points and ultimately produce dense point clouds, which shows that it is capable of reconstructing a full point cloud frame from a single compact but high dimensional descriptor. Our proposed network demonstrates performance that is comparable to the state-of-the-art approaches. With the benefit of the LPD-AE, many computationally complex tasks that rely directly on point clouds can be effortlessly conducted on latent space with lower memory costs, such as relocalization, loop closure detection, and map compression reconstruction. Comprehensive validations on Oxford RobotCar dataset, KITTI dataset, and our freshly collected dataset, which contains multiple trials of repeated routes in different weather and at different times, manifest its potency for real robotic and self-driving implementation. The source code is available at https://github.com/Suoivy/LPD-AE .

A Novel Motion Detection Method Using 3D Discrete Wavelet Transform

Recently, there has been a considerable attention given to the motion detection problem due to the explosive growth of its applications in video analysis and surveillance systems. While the previous approaches can produce good results, an accurate detection of motion remains a challenging task due to the difficulties raised by illumination variations, occlusion, camouflage, burst physical motion, dynamic texture, and environmental changes such as those on climate changes, sunlight changes during a day, etc. In this paper, we propose a novel per-pixel motion descriptor for both motion detection and dynamic texture segmentation which outperforms the current methods in the literature particularly in severe scenarios. The proposed descriptor is based on two complementary three-dimensional-discrete wavelet transform (3D-DWT) and three-dimensional wavelet leader. In this approach, a feature vector is extracted for each pixel by applying a novel three dimensional wavelet-based motion descriptor. Then, the extracted features are clustered by a clustering method such as well-known k-means algorithm or Gaussian Mixture Model (GMM). The experimental results demonstrate the effectiveness of our proposed method compared to the other motion detection approaches from the literature. The application of the proposed method and additional experimental results for the different datasets are available at (http://dspl.ce.sharif.edu/motiondetector.html).

3D shape classification with NNLS coding and optimal projections technique

Using graph-based encoding techniques and well-known shape descriptors a framework is presented here which is checked for 3D shape classification performance. A data-driven feature extraction procedure, taking the form of a simple projection operator, is in the core of our framework. Provided a shape database, a graph encapsulating the structural relationships among all the available shapes is first constructed and then is employed to define low-dimensional sparse projections. The weights in the graph, reflecting the database structure, are calculated so as to approximate each shape as a sparse linear combination of the remaining dataset objects. NNLS (nonnegative least squares) coding method is employed and fully exploited in this stage. Sparse coding with L2graph is also included in the framework. By way of solving a generalized eigenanalysis problem, a linear matrix operator is designed by means of optimal projections that will act as the feature extractor. A trained SVM (support vector machine) in the final stage of our framework makes the class prediction. Two popular, inherently high dimensional descriptors, namely ShapeDNA and Global Point Signature (GPS), with a modification emphasizing the smaller eigenvalues are employed in our experimentations with SHREC10, SHREC11 and SCHREC15 datasets. Classification results are very promising and outperform state of the art methods, providing evidence about the highly discriminative nature of the 3D shape representation produced in the proposed multistage framework.

Chemometric modeling to predict air half-life of persistent organic pollutants (POPs)

We have reported here a quantitative structure-property relationship (QSPR) model for prediction of air half-life of organic chemicals using a dataset of 302 diverse organic chemicals employing only two-dimensional descriptors with definite physicochemical meaning in order to avoid the computational complexity for higher dimensional molecular descriptors. The developed model was rigorously validated using the internationally accepted internal and external validation metrics. The final partial least squares (PLS) regression model obtained at three latent variables comprises six simple and interpretable 2D descriptors. The simple and highly robust model with good quality of predictions explains 66% for the variance of the training set (R2) (64% in terms of LOO variance (Q2)) and 76% for test set variance (R2pred) (prediction quality). This model might be applicable for data gap filling for determination of POPs in the environment, in case of new or untested chemicals falling within the applicability domain of the model. In general, the model indicates that the air half-life of organic chemicals increases with presence of H-bond acceptor atoms, number of halogen atoms and presence of the R—CH-X fragment and lipophilicity, and decreases with presence of a number of halogens on ring C(sp3) (substitution of halogen atoms on a ring).