Effective Computational Framework Research Articles

AbInitio Self-Trapped Excitons.

We propose a new formalism and an effective computational framework to study self-trapped excitons (STEs) in insulators and semiconductors from first principles. Using the many-body Bethe-Salpeter equation in combination with perturbation theory, we are able to obtain the mode- and momentum-resolved exciton-phonon coupling matrix element in a perturbative scheme and explicitly solve the real space localization of the electron (hole), as well as the lattice distortion. Further, this method allows us to compute the STE potential energy surface and evaluate the STE formation energy and Stokes shift. We demonstrate our approach using two-dimensional magnetic semiconductor chromium trihalides and a wide-gap insulator BeO, the latter of which features dark excitons, and make predictions of their Stokes shift and coherent phonon generation which we hope will spark future experiments such as photoluminescence and transient absorption studies.

CapsEnhancer: An Effective Computational Framework for Identifying Enhancers Based on Chaos Game Representation and Capsule Network.

Enhancers are a class of noncoding DNA, serving as crucial regulatory elements in governing gene expression by binding to transcription factors. The identification of enhancers holds paramount importance in the field of biology. However, traditional experimental methods for enhancer identification demand substantial human and material resources. Consequently, there is a growing interest in employing computational methods for enhancer prediction. In this study, we propose a two-stage framework based on deep learning, termed CapsEnhancer, for the identification of enhancers and their strengths. CapsEnhancer utilizes chaos game representation to encode DNA sequences into unique images and employs a capsule network to extract local and global features from sequence "images". Experimental results demonstrate that CapsEnhancer achieves state-of-the-art performance in both stages. In the first and second stages, the accuracy surpasses the previous best methods by 8 and 3.5%, reaching accuracies of 94.5 and 95%, respectively. Notably, this study represents the pioneering application of computer vision methods to enhancer identification tasks. Our work not only contributes novel insights to enhancer identification but also provides a fresh perspective for other biological sequence analysis tasks.

Predicting Cardiotoxicity of Molecules Using Attention-Based Graph Neural Networks.

In drug discovery, the search for new and effective medications is often hindered by concerns about toxicity. Numerous promising molecules fail to pass the later phases of drug development due to strict toxicity assessments. This challenge significantly increases the cost, time, and human effort needed to discover new therapeutic molecules. Additionally, a considerable number of drugs already on the market have been withdrawn or re-evaluated because of their unwanted side effects. Among the various types of toxicity, drug-induced heart damage is a severe adverse effect commonly associated with several medications, especially those used in cancer treatments. Although a number of computational approaches have been proposed to identify the cardiotoxicity of molecules, the performance and interpretability of the existing approaches are limited. In our study, we proposed a more effective computational framework to predict the cardiotoxicity of molecules using an attention-based graph neural network. Experimental results indicated that the proposed framework outperformed the other methods. The stability of the model was also confirmed by our experiments. To assist researchers in evaluating the cardiotoxicity of molecules, we have developed an easy-to-use online web server that incorporates our model.

Research Model Analysis of Muscle Power and Measuring Methods

Momentum is one of the most common sources of power for collecting. This article is important in the development of human power as it is a motivational investigation into the wood chipping process. An effective computational framework was established for the following variables: the power required to accelerate the spinning wheel, arterial pressure rise, and time needed for achieving spinning wheel speed. Five people with different physical characteristics pedaled. The energy production unit and output necessary for every individual are to be evaluated. The method of cycling and engaging the power transmitter can be considered a different subject of inquiry in this study. A separate experimental strategy was devised to measure human energy, and many factors were identified for further investigation for the first time, an effort was made to calculate the amount of human power necessary to operate any machine. The originality of the present study is regarded as of the utmost significance since it is transferable to any mathematical framework developed.

DeepAFP: An effective computational framework for identifying antifungal peptides based on deep learning.

Fungal infections have become a significant global health issue, affecting millions worldwide. Antifungal peptides (AFPs) have emerged as a promising alternative to conventional antifungal drugs due to their low toxicity and low propensity for inducing resistance. In this study, we developed a deep learning-based framework called DeepAFP to efficiently identify AFPs. DeepAFP fully leverages and mines composition information, evolutionary information, and physicochemical properties of peptides by employing combined kernels from multiple branches of convolutional neural network with bi-directional long short-term memory layers. In addition, DeepAFP integrates a transfer learning strategy to obtain efficient representations of peptides for improving model performance. DeepAFP demonstrates strong predictive ability on carefully curated datasets, yielding an accuracy of 93.29% and an F1-score of 93.45% on the DeepAFP-Main dataset. The experimental results show that DeepAFP outperforms existing AFP prediction tools, achieving state-of-the-art performance. Finally, we provide a downloadable AFP prediction tool to meet the demands of large-scale prediction and facilitate the usage of our framework by the public or other researchers. Our framework can accurately identify AFPs in a short time without requiring significant human and material resources, and hence can accelerate the development of AFPs as well as contribute to the treatment of fungal infections. Furthermore, our method can provide new perspectives for other biological sequence analysis tasks.

A Wavelet-Based Computational Framework for a Block-Structured Markov Chain with a Continuous Phase Variable

We consider the computing issues of the steady probabilities for block-structured discrete-time Markov chains that are of upper-Hessenberg or lower-Hessenberg transition kernels with a continuous phase set. An effective computational framework is proposed based on the wavelet transform, which extends and modifies the arguments in the literature for quasi-birth-death (QBD) processes. A numerical procedure is developed for computing the steady probabilities based on the fast discrete wavelet transform, and several examples are presented to illustrate its effectiveness.

SMAP: Similarity-based matrix factorization framework for inferring miRNA-disease association

Background:Based on increasing evidence, microRNAs (miRNAs) play significant roles in various complex human diseases. Therefore, identifying the disease-related miRNAs could help to provide a stepping stone to help understand the disease pathogenesis mechanism. However, owing to the cost and complexity of clinical methods, the development of novel computational frameworks that identify miRNA-disease associations could be a great alternative. Herein, we sought to present a simple yet effective computational framework (SMAP) for identifying miRNA-disease associations by applying the recommended algorithm with miRNA and disease similarity constraints. The comprehensive and accurate similarity values were measured based on miRNA functional similarity, disease semantic similarity, and Gaussian interaction profile kernel similarity. SMAP not only exploited known miRNA-disease associations for the matrix factorization model, but also utilized the integrated similarities for miRNAs and diseases. As a result, SMAP had AUCs of 0.9227 and 0.8952 in the frameworks of global and local leave-one-out cross validation, respectively. In addition, independent case studies on two major human cancers clearly verified the comparable performance of SMAP. Overall, SMAP could serve as an effective guide to elucidating pathogenesis and etiology of human diseases and deciphering the potential disease biomarkers.

On damage behavior and stability of composite T-shaped stiffened panels under compression after impact considering impact locations

Impact locations play a critical role in determining the residual performance of the composite stiffened panel, since the local damage behavior at different locations is not identical. The objective of this paper is to systematically investigate the influence of impact locations on damage behavior, stability and residual compression strength of T-shaped stiffened CFRP panels. Low-velocity impact tests are carried out by impacting at the skin center or the flange tip. The damage patterns and post-buckling behavior are monitored by ultrasonic C-scan and digital image correlation. An effective computational framework with explicit formulations considering complicated damage mechanisms is proposed to investigate the details of damage behavior. The skin center impact causes severe fiber breakage and matrix squeezing, while the flange tip impact induces complex damage patterns including ply splitting, fiber breakage, matrix cracks, as well as extensive skin–rib interfacial delamination. Compression after impact results indicate that impact damage at the skin center has little effect on the residual performance of the structure, and the compression failure load is only reduced by 2.02%, from 276.8 kN to 271.2 kN. However, impact damage at the flange tip weakens the supporting effect of the stiffener, leading to early buckling of the local skin. The propagation of the interfacial delamination at the damaged flange tip directly triggers the premature compression failure, resulting in an 18.79% decrease in the failure load, 224.8 kN. Numerical predictions correlate well with experimental data. The reported results may be useful in structural design and maintenance.

PPMS: A framework to Profile Primary MicroRNAs from Single-cell RNA-sequencing datasets.

Single-cell/nuclei RNA-sequencing (scRNA-seq) technologies can simultaneously quantify gene expression in thousands of cells across the genome. However, the majority of the noncoding RNAs, such as microRNAs (miRNAs), cannot currently be profiled at the same scale. MiRNAs are a class of small noncoding RNAs and play an important role in gene regulation. MiRNAs originate from the processing of primary transcripts, known as primary-microRNAs (pri-miRNAs). The pri-miRNA transcripts, independent of their cognate miRNAs, can also function as long noncoding RNAs, code for micropeptides or even interact with DNA, acting like enhancers. Therefore, it is apparent that the significance of scRNA-seq pri-miRNA profiling expands beyond using pri-miRNA as proxies of mature miRNAs. However, there are no computational methods that allow profiling and quantification of pri-miRNAs at the single-cell-type resolution. We have developed a simple yet effective computational framework to profile pri-MiRNAs from single-cell RNA-sequencing datasets (PPMS). Based on user input, PPMS can profile pri-miRNAs at cell-type resolution. PPMS can be applied to both newly produced and publicly available datasets obtained via single cell or single-nuclei RNA-seq. It allows users to (i) investigate the distribution of pri-miRNAs across cell types and cell states and (ii) establish a relationship between the number of cells/reads sequenced and the detection of pri-miRNAs. Here, to demonstrate its efficacy, we have applied PPMS to publicly available scRNA-seq data generated from (i) individual chambers (ventricles and atria) of the human heart, (ii) human pluripotent stem cells during their differentiation into cardiomyocytes (the heart beating cells) and (iii) hiPSCs-derived cardiomyocytes infected with severe acute respiratory syndromecoronavirus 2.

Improving molecular property prediction through a task similarity enhanced transfer learning strategy

SummaryDeeply understanding the properties (e.g., chemical or biological characteristics) of small molecules plays an essential role in drug development. A large number of molecular property datasets have been rapidly accumulated in recent years. However, most of these datasets contain only a limited amount of data, which hinders deep learning methods from making accurate predictions of the corresponding molecular properties. In this work, we propose a transfer learning strategy to alleviate such a data scarcity problem by exploiting the similarity between molecular property prediction tasks. We introduce an effective and interpretable computational framework, named MoTSE (Molecular Tasks Similarity Estimator), to provide an accurate estimation of task similarity. Comprehensive tests demonstrated that the task similarity derived from MoTSE can serve as useful guidance to improve the prediction performance of transfer learning on molecular properties. We also showed that MoTSE can capture the intrinsic relationships between molecular properties and provide meaningful interpretability for the derived similarity.

On the principles of Parsimony and Self-consistency for the emergence of intelligence

Ten years into the revival of deep networks and artificial intelligence, we propose a theoretical framework that sheds light on understanding deep networks within a bigger picture of intelligence in general. We introduce two fundamental principles, Parsimony and Self-consistency, which address two fundamental questions regarding intelligence: what to learn and how to learn, respectively. We believe the two principles serve as the cornerstone for the emergence of intelligence, artificial or natural. While they have rich classical roots, we argue that they can be stated anew in entirely measurable and computable ways. More specifically, the two principles lead to an effective and efficient computational framework, compressive closed-loop transcription, which unifies and explains the evolution of modern deep networks and most practices of artificial intelligence. While we use mainly visual data modeling as an example, we believe the two principles will unify understanding of broad families of autonomous intelligent systems and provide a framework for understanding the brain.

A comprehensive analysis of damage behaviors of composite sandwich structures under localized impact

Sandwich structure with honeycomb core and CFRP facesheets is increasingly used in the high-performance fields due to the high specific stiffness and strength. However, it is susceptible to localized low-velocity impacts such as tool drop during maintenance, and thus, understanding the damage behavior is important for a safety design and optimization. The objective of this article is to investigate the influence of the structural parameters and impact energy on the damage behaviors of the composite honeycomb sandwich panels subjected to localized low-velocity impact. An effective computational framework considering complicate damage mechanisms is proposed to predict the failure patterns, load history, and energy absorption. Comparisons with the experimental data show that the error of predictions is within 4% and thus the framework is validated. Then, a comprehensive parametric study is conducted to explore the effects of wall thickness and height of honeycomb, impact energy, and stacking sequences of the laminated facesheets on the impact resistance, damage behaviors, and energy absorption of the sandwich structure. The reported results may be useful in practice.

Cooperatively interface role of surface atoms and aqueous media on single atom catalytic property for H2O2 synthesis

Direct electrosynthesis of hydrogen peroxide (H2O2) from H2 and O2 is a promising alternative to currently industrial Riedl-Pfleiderer route. Utilizing a combination of density functional theory (DFT) and ab-initio-molecular dynamic simulation (AIMD), we presented an effective computational framework to identify the cooperative role of surface atoms(e.g. O, N and S) and aqueous media on catalytic performance of single-atom catalysts (SACs) supported Nb2C MXenes. Computational results shown that both Ni/Nb2CN2 and Co/Nb2CS2 have low overpotentials of 0.17 V and 0.20 V, and the barrier of 0.89 eV and 0.67 eV for 2e- ORR under gas phase, respectively, while in aqueous phase, hydrogen bond framework on the surface promotes the transfer of proton, resulting in the lower 2e- ORR overpotential (0.05 V) in Co/Nb2CS2 and lower barrier (almost 0.01 eV) for rate-determining step (RDS) in Ni/Nb2CN2. Electronically, we found that the less-electronegativity N and S relative to O more benefit to mediate the activation degree of O2 on SACs and thereby improve catalytic selectivity. Thus, it is concluded that both surface atom and aqueous medium synergistically promote catalytic property for H2O2 synthesis.

Synergistic Effect of Coordination Fields and Hydrosolvents on the Single-Atom Catalytic Property in H2O2 Synthesis: A Density Functional Theory Study

The design of high-performance single-atom catalysts (SACs) for two-electron oxygen reduction reaction (2e– ORR) is of great importance for the large-scale commercial application of H2O2 electrosynthesis. Herein, utilizing the combination of density functional theory (DFT) calculation and ab initio molecular dynamic (AIMD) simulation, an effective computational framework was built to identify the critical role of coordination fields and hydrosolvents in the 2e– ORR catalytic properties of SAC-based two substrates [pristine γ-type graphyne (γ-GY) and B-doped γ-GY]. The screening calculations of 168 TM@(B-doped) γ-GY (TM = transition metal) revealed that both Co@B1-2-γ-GY and Pd@B3-2-γ-GY feature outstanding catalytic properties with low overpotentials (0.06, 0.02 V) and excellent stability. The electronic structure results uncovered that viewing from the coordination field, electron-deficient boron atoms can manipulate the electron back-donation effect (d → π2p*) of the TM–O bond to mediate the H2O2 property. Solvent effect calculations revealed that the hydrogen-bond (H-bond) framework plays an important role in the high selectivity of H2O2 by facilitating the H transfer from water (>99.9 and 98.6% for Co@B1-2-γ-GY and Pd@B3-2-γ-GY), shedding light on the higher accuracy of an explicit solvent than that of an implicit one. This work will provide one way to better design high-selectivity SACs for H2O2 synthesis effectively.

A Computational Framework for Pre-Interventional Planning of Peripheral Arteriovenous Malformations

PurposePeripheral arteriovenous malformations (pAVMs) are congenital lesions characterised by abnormal high-flow, low-resistance vascular connections—the so-called nidus—between arteries and veins. The mainstay treatment typically involves the embolisation of the nidus, however the complexity of pAVMs often leads to uncertain outcomes. This study aims at developing a simple, yet effective computational framework to aid the clinical decision making around the treatment of pAVMs using routinely acquired clinical data.MethodsA computational model was developed to simulate the pre-, intra-, and post-intervention haemodynamics of a patient-specific pAVM. A porous medium of varying permeability was employed to simulate the sclerosant effect on the nidus haemodynamics. Results were compared against clinical data (digital subtraction angiography, DSA, images) and experimental flow-visualization results in a 3D-printed phantom of the same pAVM.ResultsThe computational model allowed the simulation of the pAVM haemodynamics and the sclerotherapy-induced changes at different interventional stages. The predicted inlet flow rates closely matched the DSA-derived data, although the post-intervention one was overestimated, probably due to vascular system adaptations not accounted for numerically. The nidus embolization was successfully captured by varying the nidus permeability and increasing its hydraulic resistance from 0.330 to 3970 mmHg s ml−1. The nidus flow rate decreased from 71% of the inlet flow rate pre-intervention to 1%: the flow completely bypassed the nidus post-intervention confirming the success of the procedure.ConclusionThe study demonstrates that the haemodynamic effects of the embolisation procedure can be simulated from routinely acquired clinical data via a porous medium with varying permeability as evidenced by the good qualitative agreement between numerical predictions and both in vivo and in vitro data. It provides a fundamental building block towards a computational treatment-planning framework for AVM embolisation.

Analysis of thermal effect on buckling of imperfect FG composite plates by adaptive XIGA

This paper presents an effective computational framework based on adaptive extended isogeometric analysis (XIGA) for analyzing thermal buckling stability behavior of functional graded plates (FGPs) with flaws (e.g., holes, cracks). In this model, the locally refined non-uniform rational B-splines (LR NURBS), which can exactly represent the structure geometry and carry out the local refinement of the mesh, are used as the basis functions; and the first-order shear deformation theory is adopted to describe the kinematics of plates. According to the enhanced stresses on the mid-plane under the first critical buckling mode, a posteriori error estimator is evaluated. In addition, based on the posteriori error estimator, the meshes are automatically locally refined through refining the basis functions with large error. The obtained results illustrate the accuracy and reliability of the developed method, and show that the adaptive local refinement has better convergence ratio and higher computational efficiency than those of the uniform global refinement. Addressing the influence of some numerical aspects such as gradient index, direction and location of crack, and hole size on the critical thermal buckling temperature is also given.

Atomistic Insights Into the Degradation of Inorganic Halide Perovskite CsPbI3: A Reactive Force Field Molecular Dynamics Study.

Halide perovskites make efficient solar cells but suffer from several stability issues. The characterization of these degradation processes is challenging because of the limited spatiotemporal resolution in experiments and the absence of efficient computational methods to study these reactive processes. Here, we present the first reactive force field for molecular dynamics simulations of the phase instability and the defect-induced degradation in CsPbI3. We find that the phase transitions are driven by the anharmonic fluctuations of the atoms in the perovskite lattice. At low temperatures, the Cs cations tend to move away from their preferential positions, resulting in worse contacts with the surrounding metal halide framework which initiates the conversion to a nonperovskite phase. Moreover, our simulations of defective structures reveal that, although both iodine vacancies and interstitials are mobile in the perovskite lattice, the vacancies have a detrimental effect on the stability, leading to the decomposition of perovskites to PbI2.

An isogeometric approach to topological optimization design of auxetic composites with tri-material micro-architectures

Auxetic composites, which can be viewed as a type of rationally man-made meta-composites with Negative Poisson’s Ratios (NPRs), have received widespread discussions and applications because of the promising, multifunctional and advanced properties. Micro-architectures work as building blocks of unit cells to make up composites, which possess complex topologies of auxetic multi-material. However, it should be noted that a computational design approach with the reasonability and powerful capability of optimizing auxetic composites cells-by-cells is still in need, particularly for composites with three or more materials. In this study, the main focus is to develop an effective and efficient computational design framework for auxetic composites with the tri-material using Isogeometric Topology Optimization (ITO) method, in which a tri-material topology representation model is constructed by non-uniform rationally B-splines (NURBS) to effectively represent micro-structural topology with clear and smooth boundaries. Numerical examples are discussed to present the effectiveness of the computational design framework, where a series of novel and interesting auxetic micro-architectures with the re-entrant or chiral deformation mechanisms can be found. The significance of the current work contributes to developing a platform that enables a scientific design of auxetic composites, rather than heavily depending on human intuition.

Feature Selection Methods for Protein Biomarker Discovery from Proteomics or Multiomics Data

Untargeted mass spectrometry (MS)-based proteomics provides a powerful platform for protein biomarker discovery, but clinical translation depends on the selection of a small number of proteins for downstream verification and validation. Due to the small sample size of typical discovery studies, protein markers identified from discovery data may not be generalizable to independent datasets. In addition, a good protein marker identified using a discovery platform may be difficult to implement in verification and validation platforms. Moreover, although multiomics characterization is being increasingly used in discovery cohort studies, there is no existing method for multiomics-facilitated protein biomarker selection. Here, we present ProMS, a computational algorithm for protein marker selection. The algorithm is based on the hypothesis that a phenotype is characterized by a few underlying biological functions, each manifested by a group of coexpressed proteins. A weighted k-medoids clustering algorithm is applied to all univariately informative proteins to identify both coexpressed protein clusters and a representative protein for each cluster as markers. In two clinically important classification problems, ProMS shows superior performance compared with existing feature selection methods. ProMS can be extended to the multiomics setting (ProMS_mo) through a constrained weighted k-medoids clustering algorithm, and the protein panels selected by ProMS_mo show improved performance on independent test data compared with ProMS. In addition to superior performance, ProMS and ProMS_mo also have two unique strengths. First, the feature clusters enable functional interpretation of the selected protein markers. Second, the feature clusters provide an opportunity to select replacement protein markers, facilitating a robust transition to the verification and validation platforms. In summary, this study provides a unified and effective computational framework for selecting protein biomarkers using proteomics or multiomics data. The software implementation is publicly available at https://github.com/bzhanglab/proms.

A machine learning framework to trace tumor tissue-of-origin of 13 types of cancer based on DNA somatic mutation

Carcinoma of unknown primary (CUP), defined as metastatic cancers with unknown cancer origin, occurs in 3‐5 per 100 cancer patients in the United States. Heterogeneity and metastasis of cancer brings great difficulties to the follow-up diagnosis and treatment for CUP. To find the tissue-of-origin (TOO) of the CUP, multiple methods have been raised. However, the accuracies for computed tomography (CT) and positron emission tomography (PET) to identify TOO were 20%–27% and 24%–40% respectively, which were not enough for determining targeted therapies. In this study, we provide a machine learning framework to trace tumor tissue origin by using gene length-normalized somatic mutation sequencing data. Somatic mutation data was downloaded from the Data Portal (Release 28) of the International Cancer Genome Consortium (ICGC), and 4909 samples for 13 cancers was used to identify primary site of cancers. Optimal results were obtained based on a 600-gene set by using the random forest algorithm with 10-fold cross-validation, and the average accuracy and F1-score were 0.8822 and 0.8886 respectively across 13 types of cancer. In conclusion, we provide an effective computational framework to infer cancer tissue-of-origin by combining DNA sequencing and machine learning techniques, which is promising in assisting clinical diagnosis of cancers.