From Digital Human Modeling to Human Digital Twin: Framework and Perspectives in Human Factors

The human digital twin(HDT)emerges as a promising human-centric technology in Industry 5.0,but challenges remain in human modeling and simulation.Digital human modeling(DHM)provides solutions for modeling and simulating human physical and cognitive aspects to support ergonomic analysis.However,it has limitations in real-time data usage,personalized services,and timely interaction.The emerging HDT concept offers new possibilities by integrating multi-source data and artificial intelligence for continuous monitoring and assessment.Hence,this paper reviews the evolution from DHM to HDT and proposes a unified HDT framework from a human factors perspective.The framework comprises the physical twin,the virtual twin,and the linkage between these two.The virtual twin integrates human modeling and AI engines to enable model-data-hybrid-enabled simulation.HDT can potentially upgrade traditional ergonomic methods to intelligent services through real-time analysis,timely feedback,and bidirectional interactions.Finally,the future perspectives of HDT for industrial applications as well as technical and social challenges are discussed.In general,this study outlines a human factors perspective on HDT for the first time,which is useful for cross-disciplinary research and human factors innovation to enhance the development of HDT in industry.

Integrated analysis of human influenza A(H1N1)virus infectionrelated genes to construct a suitable diagnostic model

The genome characteristics and structural functions of coding proteins correlate with the genetic diversity of the H1N1 virus,which aids in the understanding of its underlying pathogenic mechanism.In this study,analyses of the characteristic of the H1N1 virus infection-related genes,their biological functions,and infection-related reversal drugs were performed.Additionally,we used multi-dimensional bioinformatics analysis to identify the key genes and then used these to construct a diagnostic model for the H1N1 virus infection.There was a total of 169 differently expressed genes in the samples between 21 h before infection and 77 h after infection.They were used during the protein-protein interaction(PPI)analysis,and we obtained a total of 1725 interacting genes.Then,we performed a weighted gene co-expression network analysis(WGCNA)on these genes,and we identified three modules that showed significant potential for the diagnosis of the H1N1 virus infection.These modules contained 60 genes,and they were used to construct this diagnostic model,which showed an effective prediction value.Besides,these 60 genes were involved in the biological functions of this infectious virus,like the cellular response to type I interferon and in the negative regulation of the viral life cycle.However,20 genes showed an upregulated expression as the infection progressed.Other 36 upregulated genes were used to examine the relationship between genes,human influenza A virus,and infection-related reversal drugs.This study revealed numerous important reversal drug molecules on the H1N1 virus.They included rimantadine,interferons,and shikimic acid.Our study provided a novel method to analyze the characteristic of different genes and explore their corresponding biological function during the infection caused by the H1N1 virus.This diagnostic model,which comprises 60 genes,shows that a significant predictive value can be the potential biomarker for the diagnosis of the H1N1 virus infection.

Proteome-Wide Profiling of the Covalent-Druggable Cysteines with a Structure-Based Deep Graph Learning Network

Covalent ligands have attracted increasing attention due to their unique advantages,such as long residence time,high selectivity,and strong binding affinity.They also show promise for targets where previous efforts to identify noncovalent small molecule inhibitors have failed.However,our limited knowledge of covalent binding sites has hindered the discovery of novel ligands.Therefore,developing in silico methods to identify covalent binding sites is highly desirable.Here,we propose DeepCoSI,the first structure-based deep graph learning model to identify ligandable covalent sites in the protein.By integrating the characterization of the binding pocket and the interactions between each cysteine and the surrounding environment,DeepCoSI achieves state-of-the-art predictive performances.The validation on two external test sets which mimic the real application scenarios shows that DeepCosI has strong ability to distinguish ligandable sites from the others.Finally,we profiled the entire set of protein structures in the RCSB Protein Data Bank(PDB)with DeepCoSI to evaluate the ligandability of each cysteine for covalent ligand design,and made the predicted data publicly available on website.