Service Code Research Articles

Optimizing Public Services through Industrial Work Practice Management Applications Using Extreme Programming Methods

Industrial work practice or industrial Internship is applying knowledge and skills to solve real problems. It becomes one of the efforts to improve the quality of human resources in an agency or company. Managing the industrial Internship at PT Artha Abadi is still being carried out using physical documents. This allows errors to occur in recording Industrial Internship data and documents. This study aimed to overcome existing problems so that the management of Industrial Internship activities at PT Artha Abadi is more systematic, organized and efficient. The method used in creating this system is the Extreme Programming method which aims to increase effectiveness and efficiency in the software development process. The Industrial Internship Management Application is built using the Laravel framework and MySQL as its database. The tools used to create this system are XAMPP as a database server and Visual Studio Code for a text editor, in system testing using Black Box Testing to test the software's functionality. Testing shows that the Industrial Internship Management Application does not experience system errors or bugs. The results of this study show that the application is built according to the main objectives and is ready to be implemented so that the Industrial Internship at PT Artha Abadi runs more systematically, organized, and efficiently.

PPI-hotspotID for detecting protein–protein interaction hot spots from the free protein structure

Experimental detection of residues critical for protein–protein interactions (PPI) is a time-consuming, costly, and labor-intensive process. Hence, high-throughput PPI-hot spot prediction methods have been developed, but they have been validated using relatively small datasets, which may compromise their predictive reliability. Here, we introduce PPI-hotspotID, a novel method for identifying PPI-hot spots using the free protein structure, and validated it on the largest collection of experimentally confirmed PPI-hot spots to date. We explored the possibility of detecting PPI-hot spots using (i) FTMap in the PPI mode, which identifies hot spots on protein–protein interfaces from the free protein structure, and (ii) the interface residues predicted by AlphaFold-Multimer. PPI-hotspotID yielded better performance than FTMap and SPOTONE, a webserver for predicting PPI-hot spots given the protein sequence. When combined with the AlphaFold-Multimer-predicted interface residues, PPI-hotspotID yielded better performance than either method alone. Furthermore, we experimentally verified several PPI-hotspotID-predicted PPI-hot spots of eukaryotic elongation factor 2. Notably, PPI-hotspotID can reveal PPI-hot spots not obvious from complex structures, including those in indirect contact with binding partners. PPI-hotspotID serves as a valuable tool for understanding PPI mechanisms and aiding drug design. It is available as a web server (https://ppihotspotid.limlab.dnsalias.org/) and open-source code (https://github.com/wrigjz/ppihotspotid/).

PPI-hotspotID for detecting protein-protein interaction hot spots from the free protein structure.

Experimental detection of residues critical for protein-protein interactions (PPI) is a time-consuming, costly, and labor-intensive process. Hence, high-throughput PPI-hot spot prediction methods have been developed, but they have been validated using relatively small datasets, which may compromise their predictive reliability. Here, we introduce PPI-hotspotID, a novel method for identifying PPI-hot spots using the free protein structure, and validated it on the largest collection of experimentally confirmed PPI-hot spots to date. We explored the possibility of detecting PPI-hot spots using (i) FTMap in the PPI mode, which identifies hot spots on protein-protein interfaces from the free protein structure, and (ii) the interface residues predicted by AlphaFold-Multimer. PPI-hotspotID yielded better performance than FTMap and SPOTONE, a webserver for predicting PPI-hot spots given the protein sequence. When combined with the AlphaFold-Multimer-predicted interface residues, PPI-hotspotID yielded better performance than either method alone. Furthermore, we experimentally verified several PPI-hotspotID-predicted PPI-hot spots of eukaryotic elongation factor 2. Notably, PPI-hotspotID can reveal PPI-hot spots not obvious from complex structures, including those in indirect contact with binding partners. PPI-hotspotID serves as a valuable tool for understanding PPI mechanisms and aiding drug design. It is available as a web server (https://ppihotspotid.limlab.dnsalias.org/) and open-source code (https://github.com/wrigjz/ppihotspotid/).

SAAMBE-MEM: a sequence-based method for predicting binding free energy change upon mutation in membrane protein-protein complexes.

Mutations in protein-protein interactions can affect the corresponding complexes, impacting function and potentially leading to disease. Given the abundance of membrane proteins, it is crucial to assess the impact of mutations on the binding affinity of these proteins. Although several methods exist to predict the binding free energy change due to mutations in protein-protein complexes, most require structural information of the protein complex and are primarily trained on the SKEMPI database, which is composed mainly of soluble proteins. A novel sequence-based method (SAAMBE-MEM) for predicting binding free energy changes (ΔΔG) in membrane protein-protein complexes due to mutations has been developed. This method utilized the MPAD database, which contains binding affinities for wild-type and mutant membrane protein complexes. A machine learning model was developed to predict ΔΔG by leveraging features such as amino acid indices and position-specific scoring matrices (PSSM). Through extensive dataset curation and feature extraction, SAAMBE-MEM was trained and validated using the XGBoost regression algorithm. The optimal feature set, including PSSM-related features, achieved a Pearson correlation coefficient of 0.64, outperforming existing methods trained on the SKEMPI database. Furthermore, it was demonstrated that SAAMBE-MEM performs much better when utilizing evolution-based features in contrast to physicochemical features. The method is accessible via a web server and standalone code at http://compbio.clemson.edu/SAAMBE-MEM/. The cleaned MPAD database is available at the website.

Causal inference of server- and client-side code smells in web apps evolution

ContextCode smells (CS) are symptoms of poor design and implementation choices that may lead to increased defect incidence, decreased code comprehension, and longer times to release. Web applications and systems are seldom studied, probably due to the heterogeneity of platforms (server and client-side) and languages, and to study web code smells, we need to consider CS covering that diversity. Furthermore, the literature provides little evidence for the claim that CS are a symptom of poor design, leading to future problems in web apps.ObjectiveTo study the quantitative evolution and inner relationship of CS in web apps on the server- and client-sides, and their impact on maintainability and app time-to-release (TTR).MethodWe collected and analyzed 18 server-side, and 12 client-side code smells, aka web smells, from consecutive official releases of 12 PHP typical web apps, i.e., with server- and client-code in the same code base, summing 811 releases. Additionally, we collected metrics, maintenance issues, reported bugs, and release dates. We used several methodologies to devise causality relationships among the considered irregular time series, such as Granger-causality and Information Transfer Entropy(TE) with CS from previous one to four releases (lag 1 to 4).ResultsThe CS typically evolve the same way inside their group and its possible to analyze them as groups. The CS group trends are: Server, slowly decreasing; Client-side embed, decreasing and JavaScript,increasing. Studying the relationship between CS groups we found that the "lack of code quality", measured with CS density proxies, propagates from client code to server code and JavaScript in half of the applications. We found causality relationships between CS and issues. We also found causality from CS groups to bugs in Lag 1, decreasing in the subsequent lags. The values are 15% (lag1), 10% (lag2), and then decrease. The group of client-side embed CS still impacts up to 3 releases before. In group analysis, server-side CS and JavaScript contribute more to bugs. There are causality relationships from individual CS to TTR on lag 1, decreasing on lag 2, and from all CS groups to TTR in lag1, decreasing in the other lags, except for client CS.ConclusionsThere is statistical inference between CS groups. There is also evidence of statistical inference from the CS to web applications’ issues, bugs, and TTR. Client and server-side CS contribute globally to the quality of web applications, this contribution is low, but significant. Depending on the outcome variable (issues, bugs, time-to-release), the contribution quantity from CS is between 10% and 20%.

PlasCAT: Plasmid Cloud Assembly Tool.

PlasCAT (Plasmid Cloud Assembly Tool) is an easy-to-use cloud-based bioinformatics tool that enables de novo plasmid sequence assembly from raw sequencing data. Nontechnical users can now assemble sequences from long reads and short reads without ever touching a line of code. PlasCAT uses high-performance computing servers to reduce run times on assemblies and deliver results faster. PlasCAT is freely available on the web at https://sequencing.genofab.com. The assembly pipeline source code and server code are available for download at https://bitbucket.org/genofabinc/workspace/projects/PLASCAT. Click the Cancel button to access the source code without authenticating. Web servers implemented in React.js and Python, with all major browsers supported.

Analysis of Interaction Mechanisms between Java and JavaScript Programming Languages: from LiveConnect to Graal.js

The article is devoted to the study of the mechanisms of interaction between the popular programming languages Java and JavaScript using a retrospective approach. The author consistently examines the development of tools of this kind, starting from LiveConnect technology and Java applets of the mid-1990s and ending with the interpreters of the last decade. Particular attention is paid to comparing the performance of the currently relevant tools Nashorn and Graal.js. The latter is an implementation of the JavaScript language based on the Truffle framework and Context API, which is a Polyglot context for executing all supported guest languages with the Graal JIT compiler, ensuring their optimal performance. The author gives a definition to the term "warm-up", which is recommended by the GraalVM developers to achieve the highest performance. This is a kind of pre-compilation mode, the process of repeated preliminary execution of JavaScript code by the interpreter before its actual usage in order to obtain information about the most frequently utilized fragments of the program, its functions, types and data structures, alongside with their subsequent caching and optimization. The results of testing based on a project from Oracle and the authors application made it possible to verify the results declared by the GraalVM developers and can be used to optimize and improve the performance of multilingual projects using a unified technology stack, as well as when migrating the server code base from Java to JavaScript.

Enhancing Individual Worker Risk Awareness: A Location-Based Safety Check System for Real-Time Hazard Warnings in Work-Zones

This study focuses on improving pre-emptive risk recognition and safety checks to prevent workplace accidents. It underscores improvements by addressing existing research issues, suggesting potential enhancements through system development. Work approval procedures and workers’ prior risk awareness, through the confirmation of work safety standards in physically separated work areas, are fundamental methods of preventing serious accidents and creating a safe work environment. A key factor concerning worker safety is recognizing the potential accident risk factors (or hazards) in advance through practical job hazard analysis and consequently take risk-reduction measures in case the safety standards are not met. Despite the crucial significance of pre-awareness of work risks at the majority of construction sites, tools to enhance this awareness are currently limited. Furthermore, real-time detection of work risks and the implementation of risk reduction measures are contingent upon a monitoring environment and a robust safety culture. This study proposed construction worker location-tracking technology that recognizes personal identification (ID). A safety check system based on location tracking combining personal quick response code (QR code) recognition and computer vision technology to automatically identify workers’ personal identities and track their physical location was proposed. A real-time safety check system was implemented to classify automatically whether workers have confirmed hazards and to approve a work process in high-risk workplaces by supervisors. Location-tracking technology with ID recognition performed the following two safety management functions. First, if a construction worker does not pre-check the work risk information before entering the work zone, the geofencing technology automatically classifies workers as those who are not aware of job hazards. Secondly, the safety manager or supervisor entering the on-site work zone possesses the authority to halt work if the work environment fails to meet safety standards and can issue warnings regarding risky situations. Essential functions were validated through a case study involving preliminary testing within the development system. To assess the practical application of the system, virtual simulations were conducted using recorded videos from a construction site to replicate the two essential functions of the system. The system was constructed using an Apache server and Python code, and for testing purposes, the names of the workers were randomized.

Characterization of PIK3CA mutation testing in metastatic breast adenocarcinomas: a multi-site, single institution study

Abstract Introduction Detection of somatic mutations in advanced breast cancer may inform prognosis and therapy for patients. In ER+ breast cancer, nearly 40% of tumors have activating mutations in PI3 kinase (PIK3CA) for which targeted PI3 kinase inhibitors are available. At our institution, reflex ordered testing of PIK3CA in ER-positive, HER2-negative advanced breast cancers was implemented in 2017 to standardize the molecular biomarker testing in these tumors. Methods We analyzed 219 cases of advanced breast carcinoma diagnosed from 1/1/2017 to 12/31/2022 across the Houston Methodist system. With IRB approval, records were collated through full-text searching of interpretation, comment, and tumor synoptic fields in the anatomic pathology (AP) reports for “breast” and keywords for molecular oncologic testing, including “PIK3CA” and “MOLECULAR DIAGNOSTICS”, and then manually reviewed to exclude indeterminate/undifferentiated carcinomas and carcinomas of non-breast sites. Only molecular tests performed within this timeframe were considered. PIK3CA testing, both single-gene and targeted panels, was performed in 179 cases. Turnaround time (TAT) statistics, defined as days from AP report signout to molecular signout, were available for 102 cases. Cases with TATs greater than 60 days and referral laboratory testing were excluded. Data extraction, transformation, and loading was performed in Microsoft SQL Server (SSMS 18.0) and VS Code (Python 3.6). Analysis was performed using R-4.2.1 and Microsoft Excel. Results PIK3CA mutation testing was performed in 179 of 219 (82%) cases of all specimen types. The most common reasons for lack of reflex ordered molecular testing included insufficient tissue quantity and specimen decalcification. At least one PIK3CA genomic alteration was detected in 57 of 179 (32%) of cases. The most common alterations included p.H1047R (23/179, 13%), p.E545K (18/179, 10%), and p.E542K (8/179, 4%). One case had PIK3CA amplification, and one case had co-mutations of PIK3CA (p.G106V, p.H1047R). 217/219 (99%) of patients were female; no PIK3CA alterations were identified in the 2 male patients. PIK3CA frequency was not associated with age (mean age 65.4 for PIK3CA alteration detected and 63.8 for not detected, p=0.41). Mean TAT was 16.4 days (n=102), and 56% (57/102) of cases were resulted within 2 weeks of AP reporting. 22% (40/179) of cases had PIK3CA performed at a referral laboratory, largely occurring during the COVID-19 pandemic. Conclusion A reflex ordered strategy for PIK3CA mutations in advanced breast cancers led to the prompt availability of appropriate molecular results for most patients. PIK3CA alterations were present in 32% of cases, with the most common mutations being p.H1047R, p.E545K, and p.E542K. No significant differences in PIK3CA alterations by patient age was observed. Some cases had extended TATs for molecular results, which may be due to ordering practices, the specific PIK3CA testing procedure, and test send-out. Further studies will investigate the association of PIK3CA mutations and clinical staging.

Integrating Embeddings from Multiple Protein Language Models to Improve Protein O-GlcNAc Site Prediction

O-linked β-N-acetylglucosamine (O-GlcNAc) is a distinct monosaccharide modification of serine (S) or threonine (T) residues of nucleocytoplasmic and mitochondrial proteins. O-GlcNAc modification (i.e., O-GlcNAcylation) is involved in the regulation of diverse cellular processes, including transcription, epigenetic modifications, and cell signaling. Despite the great progress in experimentally mapping O-GlcNAc sites, there is an unmet need to develop robust prediction tools that can effectively locate the presence of O-GlcNAc sites in protein sequences of interest. In this work, we performed a comprehensive evaluation of a framework for prediction of protein O-GlcNAc sites using embeddings from pre-trained protein language models. In particular, we compared the performance of three protein sequence-based large protein language models (pLMs), Ankh, ESM-2, and ProtT5, for prediction of O-GlcNAc sites and also evaluated various ensemble strategies to integrate embeddings from these protein language models. Upon investigation, the decision-level fusion approach that integrates the decisions of the three embedding models, which we call LM-OGlcNAc-Site, outperformed the models trained on these individual language models as well as other fusion approaches and other existing predictors in almost all of the parameters evaluated. The precise prediction of O-GlcNAc sites will facilitate the probing of O-GlcNAc site-specific functions of proteins in physiology and diseases. Moreover, these findings also indicate the effectiveness of combined uses of multiple protein language models in post-translational modification prediction and open exciting avenues for further research and exploration in other protein downstream tasks. LM-OGlcNAc-Site’s web server and source code are publicly available to the community.

Overcoming Problems Combination of Server and Client Side Encryption on Web Code Security Using the Slice Code Method

Web security protection that is carried out by some web developers is to protect the source code of web applications by encrypting the source code, both the source code that runs on the web server or the source code that runs on the client browser. Combining these two encryption technologies is a problem for web developers. If using server side encryption, the client side source code is not protected. However, if you use client-side encryption, the server source code becomes unreadable by the web server. This study aims to overcome the problem of the combination of these two encryption technologies, so the Slice Code method is used, which is a method that the researchers themselves have developed and used to encrypt server and client source code simultaneously and combine the two. With this method, the original source code is separated into 2 parts, then the pure client source code is encrypted using client encryption technology, then embedded into the pure server source code. The next step is to encrypt the pure server source code. The result is that web applications can run perfectly by having source code encryption from both sides, namely the server side and the client side.

Web-Based Inventory and Sales Information System: Indonesian Micro Small Medium Enterprise Case Study

Information systems have a very important role in increasing productivity values that provide effectiveness and efficiency in the business sector. Cik Cik MSME Meatshop and Frozenfood is still using a manual system for recording their inventory and sales data, which causes several problems such as human error, human fraud, and inaccurate data. To overcome the problems, a solution is provided, that is a web-based information system, focused on the inventory and sales business processes. The information system is designed using Rapid Application Development method. The tools used are XAMPP as a web server and Visual Studio Code as a code editor. The programming language to be used is PHP, CodeIgniter as a framework, and MySQL as a database storage. The results obtained from the development of this information system are this system can help Cik Cik MSME Meatshop and Frozenfood in overcoming problems related to sales and inventory become easier to manage the data so that it is more automated compared to the previous manual system.

LigBind: Identifying Binding Residues for Over 1000 Ligands with Relation-Aware Graph Neural Networks

Identifying the interactions between proteins and ligands is significant for drug discovery and design. Considering the diverse binding patterns of ligands, the ligand-specific methods are trained per ligand to predict binding residues. However, most of the existing ligand-specific methods ignore shared binding preferences among various ligands and generally only cover a limited number of ligands with a sufficient number of known binding proteins. In this study, we propose a relation-aware framework LigBind with graph-level pre-training to enhance the ligand-specific binding residue predictions for 1159 ligands, which can effectively cover the ligands with a few known binding proteins. LigBind first pre-trains a graph neural network-based feature extractor for ligand-residue pairs and relation-aware classifiers for similar ligands. Then, LigBind is fine-tuned with ligand-specific binding data, where a domain adaptive neural network is designed to automatically leverage the diversity and similarity of various ligand-binding patterns for accurate binding residue prediction. We construct ligand-specific benchmark datasets of 1159 ligands and 16 unseen ligands, which are used to evaluate the effectiveness of LigBind. The results demonstrate the LigBind’s efficacy on large-scale ligand-specific benchmark datasets, and it generalizes well to unseen ligands. LigBind also enables accurate identification of the ligand-binding residues in the main protease, papain-like protease and the RNA-dependent RNA polymerase of SARS-CoV-2. The web server and source codes of LigBind are available at http://www.csbio.sjtu.edu.cn/bioinf/LigBind/ and https://github.com/YYingXia/LigBind/ for academic use.

DEVOPS И СОВРЕМЕННАЯ СЕРВЕРНАЯ АРХИТЕКТУРА

An integral part of any multi-user service is its server part. Data storage, authorization logic, algorithm requiring
 security so that no user can find a vulnerability in the system. All this is assigned to the backend - the server
 logic of the application. Without exaggeration, thousands and millions of users around the world depend on
 security, stability, and continuous updating o f the server code. It is the responsibility o f each developer to ensure
 the maximum performance o f the above properties.
 The development o f any application depends on its architecture and the architecture o f each component. Previously,
 setting up the server environment and developing the backend o f the service part were inseparable. But
 applications are becoming more complex, system requirements are becoming more difficult to implement. Setting
 up a server environment, automating work processes and updates are highlighted in a separate area o f
 activity and the IT industry - DevOps. There are containerization tools, process automation tools, web servers
 and much more in the world. And none o f these technologies are used in isolation. For a software developer
 without appropriate DevOps skills, it is difficult to implement a competent server configuration

Auto-Kla: a novel web server to discriminate lysine lactylation sites using automated machine learning.

Recently, lysine lactylation (Kla), a novel post-translational modification (PTM), which can be stimulated by lactate, has been found to regulate gene expression and life activities. Therefore, it is imperative to accurately identify Kla sites. Currently, mass spectrometry is the fundamental method for identifying PTM sites. However, it is expensive and time-consuming to achieve this through experiments alone. Herein, we proposed a novel computational model, Auto-Kla, to quickly and accurately predict Kla sites in gastric cancer cells based on automated machine learning (AutoML). With stable and reliable performance, our model outperforms the recently published model in the 10-fold cross-validation. To investigate the generalizability and transferability of our approach, we evaluated the performance of our models trained on two other widely studied types of PTM, including phosphorylation sites in host cells infected with SARS-CoV-2 and lysine crotonylation sites in HeLa cells. The results show that our models achieve comparable or better performance than current outstanding models. We believe that this method will become a useful analytical tool for PTM prediction and provide a reference for the future development of related models. The web server and source code are available at http://tubic.org/Kla and https://github.com/tubic/Auto-Kla, respectively.

TriNet: A tri-fusion neural network for the prediction of anticancer and antimicrobial peptides

The accurate identification of anticancer peptides (ACPs) and antimicrobial peptides (AMPs) remains a computational challenge. We propose a tri-fusion neural network termed TriNet for the accurate prediction of both ACPs and AMPs. The framework first defines three kinds of features to capture the peptide information contained in serial fingerprints, sequence evolutions, and physicochemical properties, which are then fed into three parallel modules: a convolutional neural network module enhanced by channel attention, a bidirectional long short-term memory module, and an encoder module for training and final classification. To achieve a better training effect, TriNet is trained via a training approach using iterative interactions between the samples in the training and validation datasets. TriNet is tested on multiple challenging ACP and AMP datasets and exhibits significant improvements over various state-of-the-art methods. The web server and source code of TriNet are respectively available at http://liulab.top/TriNet/server and https://github.com/wanyunzh/TriNet.

Edge-Centric Programming for IoT Applications With Automatic Code Partitioning

IoT application development usually involves separate programming at the device side and server side. While separate programming style is sufficient for many simple applications, it is not suitable for many complex applications that involve complex interactions and intensive data processing. We propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EdgeProg</i> , an edge-centric programming approach to simplify IoT application programming, motivated by the increasing popularity of edge computing. With EdgeProg, users could write application logic in a centralized manner with an augmented If-This-Then-That (IFTTT) syntax and virtual sensor mechanism. The program can be processed at the edge server, which can automatically generate the actual application code and intelligently partition the code into device code and server code, for achieving the optimal latency. EdgeProg employs dynamic linking and loading to deploy the device code on a variety of IoT devices, which do not run any application-specific codes at the start. Results show that EdgeProg achieves an average reduction of 20.96%, 27.8% and 79.41% in terms of execution latency, energy consumption, and lines of code compared with state-of-the-art approaches.

Event-Based Supply Chain Network Modeling: Blockchain for Good Coffee

Blockchain and distributed ledger technology (BC/DLT) provides distributed databases with decentralized governance, tamper-proof recording, high availability and non-copyable digital assets, which have made it a natural technological basis for supply chain management. In this paper, we introduce REALISTIC, a novel event-based modeling framework for supply chain networks (SCNs) that includes production processes. It extends McCarty’s Resources-Events-Agents (REA) accounting model with secure transformations, which, across the entire SCN, guarantee that certified output resources cannot be digitally produced ex nihilo, but require certified input resources of at least the same amount as what is produced. This generalizes the no-double-spend guarantee of current BC/DLT to (digital twins of) physical resources and their production. Authenticated human or robotic Internet of Things (IoT) actors digitally sign and cryptographically commit to the veracity of real-world events on an immutable database, without having to take responsibility for their aggregate consequences. User-specifiable interpretations, corresponding to queries and analytical functions in database systems, provide auditable aggregate information computed from recorded events across the entire SCN. This includes fine-grained and trustworthy tracing of final products through multiple stages of production processes, semi-finished products, quality certifications and transportation all the way back to their raw materials. We present a case study for an end-to-end coffee supply chain that tracks fine-grained and detailed information from a farmer’s coffee cherries to retail coffee bags, involving all its actors. Our model handles product provenance; auditable sustainability, quality and trade information; production processes from parchment via green to roasted coffee; product quality tests; farmer certifications; and transportation across the entire coffee supply chain. It is based on field work involving farmers, cooperatives, processors, traders, importers, and a major roasting company stretching from Colombia to Scandinavia. Its REALISTIC-based modeling is the foundation for the design of our prototype implementation, which includes Ethereum blockchain code, RDBMS-based server code and a web app client. Their source code is publicly available on GitHub.

Cross-attention PHV: Prediction of human and virus protein-protein interactions using cross-attention–based neural networks

Viral infections represent a major health concern worldwide. The alarming rate at which SARS-CoV-2 spreads, for example, led to a worldwide pandemic. Viruses incorporate genetic material into the host genome to hijack host cell functions such as the cell cycle and apoptosis. In these viral processes, protein–protein interactions (PPIs) play critical roles. Therefore, the identification of PPIs between humans and viruses is crucial for understanding the infection mechanism and host immune responses to viral infections and for discovering effective drugs. Experimental methods including mass spectrometry-based proteomics and yeast two-hybrid assays are widely used to identify human-virus PPIs, but these experimental methods are time-consuming, expensive, and laborious. To overcome this problem, we developed a novel computational predictor, named cross-attention PHV, by implementing two key technologies of the cross-attention mechanism and a one-dimensional convolutional neural network (1D-CNN). The cross-attention mechanisms were very effective in enhancing prediction and generalization abilities. Application of 1D-CNN to the word2vec-generated feature matrices reduced computational costs, thus extending the allowable length of protein sequences to 9000 amino acid residues. Cross-attention PHV outperformed existing state-of-the-art models using a benchmark dataset and accurately predicted PPIs for unknown viruses. Cross-attention PHV also predicted human–SARS-CoV-2 PPIs with area under the curve values >0.95. The Cross-attention PHV web server and source codes are freely available at https://kurata35.bio.kyutech.ac.jp/Cross-attention_PHV/ and https://github.com/kuratahiroyuki/Cross-Attention_PHV, respectively.

Improved Protein Structure Prediction Using a New Multi-Scale Network and Homologous Templates.

The accuracy of de novo protein structure prediction has been improved considerably in recent years, mostly due to the introduction of deep learning techniques. In this work, trRosettaX, an improved version of trRosetta for protein structure prediction is presented. The major improvement over trRosetta consists of two folds. The first is the application of a new multi‐scale network, i.e., Res2Net, for improved prediction of inter‐residue geometries, including distance and orientations. The second is an attention‐based module to exploit multiple homologous templates to increase the accuracy further. Compared with trRosetta, trRosettaX improves the contact precision by 6% and 8% on the free modeling targets of CASP13 and CASP14, respectively. A preliminary version of trRosettaX is ranked as one of the top server groups in CASP14's blind test. Additional benchmark test on 161 targets from CAMEO (between Jun and Sep 2020) shows that trRosettaX achieves an average TM‐score ≈0.8, outperforming the top groups in CAMEO. These data suggest the effectiveness of using the multi‐scale network and the benefit of incorporating homologous templates into the network. The trRosettaX algorithm is incorporated into the trRosetta server since Nov 2020. The web server, the training and inference codes are available at: https://yanglab.nankai.edu.cn/trRosetta/.