Interaction Information Research Articles

GIVA: Interaction-aware trajectory prediction based on GRU-Improved VGG-Attention Mechanism model for autonomous vehicles

Predicting future trajectories is crucial for autonomous vehicles, as accurate predictions enhance safety and inform subsequent decision-making and planning modules. This is however a challenging task due to the complex interactions between surrounding vehicles. Existing methods struggled to extract deep representations and often overlook spatial dependence. To address this problem, this paper introduces GIVA, an interaction-aware trajectory prediction method based on the Gated Recurrent Unit (GRU)-Improved Visual Geometry Group (VGG)-Attention Mechanism model. GIVA first encodes the historical trajectories of the target vehicle and its surrounding vehicles using a GRU Encoder. Next, an Interaction Module, which combines the Improved VGG Pooling Module and the Attention Mechanism Pooling Module, effectively captures spatial interaction features between vehicles. The Improved VGG Pooling Module extracts more detailed and effective interaction information, while the Attention Mechanism Pooling Module emphasizes the importance of surrounding vehicles for the target vehicle’s future trajectory. Lastly, the dynamic encoding feature of the target vehicle and the fused interaction feature are concatenated and input into a GRU Decoder to generate the future trajectory. Experiments on the public Next Generation Simulation (NGSIM) dataset showcase the effectiveness of GIVA compared to existing prediction approaches, demonstrating its potential for improving autonomous vehicle performance.

EGMT-CD: Edge-Guided Multimodal Transformers Change Detection from Satellite and Aerial Images

Change detection from heterogeneous satellite and aerial images plays a progressively important role in many fields, including disaster assessment, urban construction, and land use monitoring. Currently, researchers have mainly devoted their attention to change detection using homologous image pairs and achieved many remarkable results. It is sometimes necessary to use heterogeneous images for change detection in practical scenarios due to missing images, emergency situations, and cloud and fog occlusion. However, heterogeneous change detection still faces great challenges, especially using satellite and aerial images. The main challenges in satellite and aerial image change detection are related to the resolution gap and blurred edge. Previous studies used interpolation or shallow feature alignment before traditional homologous change detection methods, which ignored the high-level feature interaction and edge information. Therefore, we propose a new heterogeneous change detection model based on multimodal transformers combined with edge guidance. In order to alleviate the resolution gap between satellite and aerial images, we design an improved spatially aligned transformer (SP-T) with a sub-pixel module to align the satellite features to the same size of the aerial ones supervised by a token loss. Moreover, we introduce an edge detection branch to guide change features using the object edge with an auxiliary edge-change loss. Finally, we conduct considerable experiments to verify the effectiveness and superiority of our proposed model (EGMT-CD) on a new satellite–aerial heterogeneous change dataset, named SACD. The experiments show that our method (EGMT-CD) outperforms many previously superior change detection methods and fully demonstrates its potential in heterogeneous change detection from satellite–aerial images.

Multi-Head Self-Attention-Enhanced Prototype Network with Contrastive–Center Loss for Few-Shot Relation Extraction

Few-shot relation extraction (FSRE) constitutes a critical task in natural language processing (NLP), involving learning relationship characteristics from limited instances to enable the accurate classification of new relations. The existing research primarily concentrates on using prototype networks for FSRE and enhancing their performance by incorporating external knowledge. However, these methods disregard the potential interactions among different prototype networks, and each prototype network can only learn and infer from its limited instances, which may limit the robustness and reliability of the prototype representations. To tackle the concerns outlined above, this paper introduces a novel prototype network called SACT (multi-head self-attention and contrastive-center loss), aimed at obtaining more comprehensive and precise interaction information from other prototype networks to bolster the reliability of the prototype network. Firstly, SACT employs a multi-head self-attention mechanism for capturing interaction information among different prototypes from traditional prototype networks, reducing the noise introduced by unknown categories with a small sample through information aggregation. Furthermore, SACT introduces a new loss function, the contrastive–center loss function, aimed at tightly clustering samples from a similar relationship category in the center of the feature space while dispersing samples from different relationship categories. Through extensive experiments on FSRE datasets, this paper demonstrates the outstanding performance of SACT, providing strong evidence for the effectiveness and practicality of SACT.

Q‐EANet: Implicit social modeling for trajectory prediction via experience‐anchored queries

AbstractAccurately predicting the future trajectory and behavior of traffic participants is crucial for the maneuvers of self‐driving vehicles. Many existing works employed a learning‐based “encoder‐interactor‐decoder” structure, but they often fail to clearly articulate the relationship between module selections and real‐world interactions. As a result, these approaches tend to rely on a simplistic stacking of attention modules. To address this issue, a trajectory prediction network (Q‐EANet) is presented in this study, which integrates GRU encoders, MLPs and attention modules. By introducing a new explanatory rule, it makes a contribution to interpretable modeling, models the entire trajectory prediction process via an implicit social modeling formula. Inspired by the anchoring effect in decision psychology, the prediction task is formulated as an information query process that occurs before traffic participants make decisions. Specifically, Q‐EANet uses GRUs to encode features and utilizes attention modules to aggregates interaction information for generating the target trajectory anchors. Then, queries are introduced for further interaction. These queries, along with the trajectory anchors with added Gaussian noise, are then processed by a GRU‐based decoder. The final prediction results are obtained through a Laplace MDN. Experimental results on the several benchmarks demonstrate the effectiveness of Q‐EANet in trajectory prediction tasks. Compared to the existing works, the proposed method achieves state‐of‐the‐art performance with only simple module design. The code for this work is publicly available at https://github.com/Jctrp/socialea.

BNDCNet: Bilateral nonlocal decoupled convergence network for semantic segmentation

The perceptual scope of deep convolutional neural networks is inherently confined to a local scale due to the inherent limitations of convolution operations. This confinement subsequently hampers a comprehensive understanding of intricate scenes. In response, an innovative approach called Bilateral Non-Local Decoupled Convergence Network (BNDCNet) is proposed to facilitate contextual interaction of global information. The proposed module decouples the information from the input feature map and uses a bilateral non-local architecture for system processing. This strategy facilitates inter-pixel interaction and aggregates global information. An important aspect of our approach involves the computation of adaptive convolutional channel weights for the feature map. This innovation greatly improves the efficacy and performance of the model. The proposed method achieved high performance on the competitive scene-parsing datasets CamVid, Cityscapes, and KITTI, and thus demonstrated effectiveness and generality. Code has been released: https://github.com/Mantee0810/BNDC.

AptaDB: a comprehensive database integrating aptamer-target interactions.

Aptamers have emerged as research hotspots of the next generation due to excellent performance benefits and application potentials in pharmacology, medicine, and analytical chemistry. Despite the numerous aptamer investigations, the lack of comprehensive data integration has hindered the development of computational methods for aptamers and the reuse of aptamers. A public access database named AptaDB, derived from experimentally validated data manually collected from the literature, was hence developed, integrating comprehensive aptamer-related data, which include six key components: (i) experimentally validated aptamer-target interaction information, (ii) aptamer property information, (iii) structure information of aptamer, (iv) target information, (v) experimental activity information, and (vi) algorithmically calculated similar aptamers. AptaDB currently contains 1350 experimentally validated aptamer-target interactions, 1230 binding affinity constants, 1293 aptamer sequences, and more. Compared to other aptamer databases, it contains twice the number of entries found in available databases. The collection and integration of the above information categories is unique among available aptamer databases and provides a user-friendly interface. AptaDB will also be continuously updated as aptamer research evolves. We expect that AptaDB will become a powerful source for aptamer rational design and a valuable tool for aptamer screening in the future. For access to AptaDB, please visit http://lmmd.ecust.edu.cn/aptadb/.

Analysis of Students’ Behavior Watching iMooX Courses with Interactive Elements

Digital learning technologies are becoming increasingly important for our modern educational system. In addition to teaching methods that incorporate interactivity, these approaches benefit students’ overall learning experience and success by enhancing their attention and fostering a positive attitude towards the learning content being presented. Interactivity comes in various forms, and while a combination of distinct activities is beneficial, some are more effective at engaging students. Using digital technologies in an educational environment opens up new possibilities for students, teachers, and researchers. It provides new insights into learning behavior and enables the collection of interaction information. This data could, for example, show how often a video was paused or at what point students lost interest and left, but gaining such knowledge requires further processing. The use of visualizations that depict behavior, such as the change of attention over time, can be an effective way to present extracted information. Therefore, our research focuses on developing an application that enables us to generate various visualizations from the collected data. A single command-line input will be sufficient to create them. Furthermore, a video course was created from which we collected behavioral data. Our results aim to showcase the benefits of interactivity, and that the created figures can be used for data evaluation verifies the versatility of the generated visualizations.

4D nucleome equation predicts gene expression controlled by long-range enhancer-promoter interaction.

Recent experimental evidence strongly supports that three-dimensional (3D) long-range enhancer-promoter (E-P) interactions have important influences on gene-expression dynamics, but it is unclear how the interaction information is translated into gene expression over time (4D). To address this question, we developed a general theoretical framework (named as a 4D nucleome equation), which integrates E-P interactions on chromatin and biochemical reactions of gene transcription. With this equation, we first present the distribution of mRNA counts as a function of the E-P genomic distance and then reveal a power-law scaling of the expression level in this distance. Interestingly, we find that long-range E-P interactions can induce bimodal and trimodal mRNA distributions. The 4D nucleome equation also allows for model selection and parameter inference. When this equation is applied to the mouse embryonic stem cell smRNA-FISH data and the E-P genomic-distance data, the predicted E-P contact probability and mRNA distribution are in good agreement with experimental results. Further statistical inference indicates that the E-P interactions prefer to modulate the mRNA level by controlling promoter activation and transcription initiation rates. Our model and results provide quantitative insights into both spatiotemporal gene-expression determinants (i.e., long-range E-P interactions) and cellular fates during development.

Libraries of Higher Education Institutions in the Prevalence of Open Science Practices

The aim of the article is to determine the state and directions for improving the implementation effectiveness of the open science concept in libraries of higher education institutions of Ukraine.
 Research methods. To achieve the set goal and solve the tasks, content analysis, systematic approach, methods of analysis and synthesis, generalization and systematization are used.
 Scientific novelty. On the basis of the formed sample of 12 libraries of Ukrainian higher education institutions (HEIs), the thematic range of electronic resources and services of institutional repositories is identified, as well as the methodical and regulatory support for their functioning is characterized. The article highlights peculiarities of implementing the basic principles of open science by libraries of Ukrainian HEIs, additionally, reveals the reserves for developing an open information space for scientists and education seekers.
 Main conclusions. Based on the analysis of Ukrainian and foreign researchers’ scientific works, the contribution of scholars to the formation of conceptual foundations of open science as a communication phenomenon, which is of key importance for the development of the digital society and knowledge society in total, is established. A number of priority tasks that need to be solved are formulated: improving the means of technological support for institutional repositories; providing libraries of HEIs with open access to subscribed resources and abstract bibliographic databases, as well as collections of digitized publications included in the State Register of National Cultural Heritage of Ukraine; providing educational materials in open access (disciplines programs, lectures notes, meetings protocols, didactic materials, etc.); elaborating and implementing a number of information, library and educational methodologies of scientific interaction with different categories of users. The significance of the resource potential of modern practices in HEIs libraries in prevalencing the achievements of open science is substantiated and summarized; the current state of implementing the open science concept in Ukrainian HEIs is determined. Prospects for further research include testing different aspects of the open access implementation; development of open science tools and services by HEIs libraries.

AAAKB: A manually curated database for tracking and predicting genes of Abdominal aortic aneurysm (AAA).

Abdominal aortic aneurysm (AAA), an extremely dangerous vascular disease with high mortality, causes massive internal bleeding due to aneurysm rupture. To boost the research on AAA, efforts should be taken to organize and link the information about AAA-related genes and their functions. Currently, most researchers screen through genetic databases manually, which is cumbersome and time-consuming. Here, we developed "AAAKB" a manually curated knowledgebase containing genes, SNPs and pathways associated with AAA. In order to facilitate researchers to further explore the mechanism network of AAA, AAAKB provides predicted genes that are potentially associated with AAA. The prediction is based on the protein interaction information of genes collected in the database, and the random forest algorithm (RF) is used to build the prediction model. Some of these predicted genes are differentially expressed in patients with AAA, and some have been reported to play a role in other cardiovascular diseases, illustrating the utility of the knowledgebase in predicting novel genes. Also, AAAKB integrates a protein interaction visualization tool to quickly determine the shortest paths between target proteins. As the first knowledgebase to provide a comprehensive catalog of AAA-related genes, AAAKB will be an ideal research platform for AAA. Database URL: http://www.lqlgroup.cn:3838/AAAKB/.

SDFormer: A shallow-to-deep feature interaction for knowledge graph embedding

Inferring missing information from current facts in a knowledge graph (KG) is the target of the link prediction task. Currently, existing methods embed the entities and relations of KG as a whole into a low-dimensional vector space. Nonetheless, they ignore the multi-level interactions (shallow interactions, deep interactions) among the finer-grained sub-features of entities and relations. To overcome these limitations, we present a shallow-to-deep feature interaction for knowledge graph embedding (SDFormer). It takes into account the interpretability of sub-feature tokens of entities and relations and learns shallow-to-deep interaction information between entities and relations at a more fine-grained level. Specifically, entity and relation vectors are decomposed into sub-features to represent multi-dimensional information. Then, a shallow-to-deep feature interaction method is designed to capture multi-level interactions between entities and relations. This process enriches the feature representation by modeling the interaction between sub-features. Finally, a 1-X scoring function is utilized to calculate the score of each knowledge triplet. The experimental results on several benchmark datasets show that SDFormer obtains competitive performance results and more efficient training efficiency on other comparative models and because of the shallow-to-deep feature interaction between entities and relations.

GTPAN: Global Target Preference Attention Network for session-based recommendation

Session-based recommendations (SBR) aim to predict the user’s next choice of items based on historical data. It is hard to access the interaction information between users and we can only analyze the user’s behavioral preferences based on the current session characteristics, which makes the recommendation task in this field quite challenging. Recently, considerable research has grown up around the theme of SBR. Previous works apply attention mechanisms and Graph Neural Networks (GNN) to achieve remarkable results. However, these methods ignore two main points: one is losing sight of the specific position of items in the session; the other is to drive global preference with local preference without considering the offset of the target preference among items in the session. This paper proposes a Global Target Preference Attention Network (GTPAN) Model with position information for the session-based recommendation. We introduce a position-aware embedding module to eliminate the effect of position missing. Considering that each item in the session is potentially rich in latent information of target preference, we use a target preference attention layer to learn the target preference vector of the session. Extensive experiments on three real datasets show that the proposed GTPAN model outperforms state-of-the-art session-based recommendation methods, especially on datasets with long sessions.

Vehicle Trajectory Prediction Based on Graph Convolutional Networks in Connected Vehicle Environment

Vehicle trajectory prediction is an important research basis for the decision making and path planning of the intelligent and connected vehicle. In the connected vehicle environment, vehicles share information and drive cooperatively, and the intelligent and connected vehicles are able to obtain more accurate and rich perception information, which provides a data basis for accurate prediction of vehicle trajectories. However, attaining accurate and effective vehicle trajectory predictions poses technical challenges due to insufficient extraction of vehicular spatial–temporal interaction features. In this paper, we propose a vehicle trajectory prediction model based on graph convolutional neural network (GCN) in a connected vehicle environment. Specifically, using the driving scene information obtained by the intelligent and connected vehicle, the spatial graph and temporal graph are constructed based on the spatial interaction coefficient (SIC) and self-attention mechanism, respectively. Then, the graph data are entered into the interaction extraction module, and the spatial interaction features and temporal interaction features are extracted separately using the graph convolutional networks, which are fused to obtain the spatial–temporal interaction information. Finally, the interaction features are learned based on the convolutional neural networks to output the future trajectory information of all vehicles in the scene by one forward operation rather than a step-by-step process. The ablation experiment results show that the method proposed in this study to model the spatiotemporal interaction among vehicles based on SIC and self-attention mechanism reduces the prediction error by 5% and 12%, respectively. The results from the model comparison experiment show that the proposed method engenders an 8% improvement in prediction accuracy over the state-of-the-art solution, providing technical and theoretical support for trajectory prediction research of intelligent and connected vehicles.

A deep learning method for drug-target affinity prediction based on sequence interaction information mining

Background A critical aspect of in silico drug discovery involves the prediction of drug-target affinity (DTA). Conducting wet lab experiments to determine affinity is both expensive and time-consuming, making it necessary to find alternative approaches. In recent years, deep learning has emerged as a promising technique for DTA prediction, leveraging the substantial computational power of modern computers. Methods We proposed a novel sequence-based approach, named KC-DTA, for predicting drug-target affinity (DTA). In this approach, we converted the target sequence into two distinct matrices, while representing the molecule compound as a graph. The proposed method utilized k-mers analysis and Cartesian product calculation to capture the interactions and evolutionary information among various residues, enabling the creation of the two matrices for target sequence. For molecule, it was represented by constructing a molecular graph where atoms serve as nodes and chemical bonds serve as edges. Subsequently, the obtained target matrices and molecule graph were utilized as inputs for convolutional neural networks (CNNs) and graph neural networks (GNNs) to extract hidden features, which were further used for the prediction of binding affinity. Results In order to evaluate the effectiveness of the proposed method, we conducted several experiments and made a comprehensive comparison with the state-of-the-art approaches using multiple evaluation metrics. The results of our experiments demonstrated that the KC-DTA method achieves high performance in predicting drug-target affinity (DTA). The findings of this research underscore the significance of the KC-DTA method as a valuable tool in the field of in silico drug discovery, offering promising opportunities for accelerating the drug development process. All the data and code are available for access on https://github.com/syc2017/KCDTA.

Predictive role of E2F6 in cancer prognosis and responses of immunotherapy

BackgroundE2F6 is a member of the E2F transcription factor family. Numerous studies have demonstrated that E2F6 is critical to cancer development and progression, but its role in cancer immunotherapy remains unclear. MethodsGenotype-Tissue Expression (GTEx) and The Cancer Genome Atlas (TCGA) databases were used to obtain RNA-seq data for cancer and normal tissues, and we utilized the cBioPortal to analyze E2F6 genomic alterations in pan-cancer. The protein localization of E2F6 was obtained using the Human Protein Atlas (HPA), and the upregulation of E2F6 expression in clinical glioblastoma multiforme (GBM) tissues was detected by Western blot analysis. The ComPPI website was used to analyze the protein interaction information of E2F6. To evaluate the role of E2F6 in pan-cancer prognosis, we used univariate Cox regression and Kaplan-Meier methods, and gene set enrichment analysis (GSEA) was utilized to identify markers associated with E2F6 expression in tumors. TIMER 2.0 was used to study E2F6-related immune cell infiltration in tumor tissues, and the correlation of E2F6 with immunotherapy biomarkers was investigated using Spearman correlation analysis. The role of E2F6 in the cell cycle was analyzed by flow cytometry, and the Cell Counting Kit-8 (CCK-8) and colony formation assays were utilized to determine the proliferative ability of cells. ResultsIn most tumor types, E2F6 was highly expressed and was a good predictor of prognosis. E2F6 was significantly related to markers of immune activation, tumor immune cell infiltration, and immune regulators. Furthermore, E2F6 knockdown significantly attenuated the proliferative ability of glioma cells. Finally, E2F6 effectively predicted anti-programmed cell death 1 (PD1) treatment response. ConclusionE2F6 is an effective biomarker that predicts the prognosis of cancer patients treated with anti-immune checkpoint therapy.

SCA-Former: transformer-like network based on stream-cross attention for medical image segmentation

Objective. Deep convolutional neural networks (CNNs) have been widely applied in medical image analysis and achieved satisfactory performances. While most CNN-based methods exhibit strong feature representation capabilities, they face challenges in encoding long-range interaction information due to the limited receptive fields. Recently, the Transformer has been proposed to alleviate this issue, but its cost is greatly enlarging the model size, which may inhibit its promotion. Approach. To take strong long-range interaction modeling ability and small model size into account simultaneously, we propose a Transformer-like block-based U-shaped network for medical image segmentation, dubbed as SCA-Former. Furthermore, we propose a novel stream-cross attention (SCA) module to enforce the network to focus on finding a balance between local and global representations by extracting multi-scale and interactive features along spatial and channel dimensions. And SCA can effectively extract channel, multi-scale spatial, and long-range information for a more comprehensive feature representation. Main results. Experimental results demonstrate that SCA-Former outperforms the current state-of-the-art (SOTA) methods on three public datasets, including GLAS, ISIC 2017 and LUNG. Significance. This work exhibits a promising method to enhance the feature representation of convolutional neural networks and improve segmentation performance.

Abstract C059: Probabilistic graph-based model uncovers druggable vulnerabilities in major solid cancers

Abstract We present A3D3a’s Molecular Vulnerability Picker (MVP), a novel probabilistic approach to determine vulnerabilities in networks of cancer interactions for the identification of new therapeutic targets. Successes in targeted therapies driven by molecular profiling have helped 47% of cancer patients achieve long-lasting remission. However, over half of patients still lack safe, effective, and long-lasting treatment options. Successes have often clustered around exploitation of major driver mechanisms that can be clearly discerned from molecular data. Heterogeneity, rarity, and complexity of these remaining cancers mean that the signal of key molecular vulnerabilities, i.e., protein targets that can be exploited for therapy, can be drowned out by noise. We developed A3D3a’s MVP to increase the signal-to-noise ratio in cancer networks of interaction and alteration information to determine vulnerabilities toward the identification of new potential drug targets. To implement A3D3a’s MVP, we constructed protein networks of major solid cancers from TCGA data that we used as a framework for information flow. We developed a novel weight-biased Markov Chain model to highlight cooperativity of weak signals arising from related regions of the protein network, emphasizing previously hidden signal within these cancer networks. To validate A3D3a’s MVP, we examined the top genes it returned in the Dependency Map and the Genomics of Drug Sensitivity in Cancer. We further validated our findings by comparing the number of genetic dependencies and drug targets recovered by the model to that recovered by the state of the art. Finally, we applied A3D3a’s MVP to identify therapeutic opportunities across cancer indications and extended our analysis to highlight well-ranked genes whose proteins contain druggable pockets and thus, could serve as new drug targets. For the 19 cancer types included in this analysis, A3D3a’s MVP returned significantly more genetic dependencies (p < 0.01) and drug targets (p < 0.001) than genes ranked by the state of the art. We demonstrate that A3D3a’s MVP is able to increase the signal of weakly altered genes and is also able to identify genuine dependencies that themselves are not mutated or altered in any way. Using A3D3a’s MVP, we identified 56 drug repurposing opportunities and 49 potential druggable targets for solid cancers and also highlight novel potential druggable targets for future exploitation and new therapeutics. In summary, A3D3a’s MVP is a novel mathematical modeling approach that increases the signal-to-noise ratio in cancer molecular data and helps uncover previously hidden molecular vulnerabilities towards new potential drug targets to address unmet patient needs. Citation Format: Stephanie T Schmidt, Ying Zhu, Li Zhao, Chunjie Jiang, Patrizio Di Micco, Costas Mitsopoulos, Andrew Futreal, Bissan Al-Lazikani. Probabilistic graph-based model uncovers druggable vulnerabilities in major solid cancers [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr C059.

Interactome‐guided machine learning modeling reveals information embedded in blood plasma protein‐protein interactions in Alzheimer’s Disease patients treated by the complex plasma fraction GRF6019

AbstractBackgroundThe relevance of plasma fraction treatment to reverse age‐related conditions was underlined by our previous pre‐clinical results. In two Phase 2 clinical trials (GRF6019‐201 and GRF6019‐202), patients with Alzheimer’s Disease (AD) were treated with the complex plasma fraction GRF6019, which was determined to be safe and well‐tolerated. Interestingly, no cognitive decline and minimal functional decline were observed. While GRF6019 treatment was associated with a quick and widespread proteomics response, lasting molecular changes were more subtle. Here, we developed new machine learning models guided by protein‐protein interaction (PPI) information to deeper understand the molecular effects of GRF6019 plasma fraction treatment.MethodMore than 7k proteins were measured in 120 plasma samples from treated subjects enrolled in the GRF6019‐201 and GRF6019‐202 trials (n = 21 and 18, respectively) using the Somascan assay. Interacting protein pairs were curated from STRING and OmniPath PPI databases. For each protein‐protein interaction, linear Support Vector Machine (SVM) models were trained to separate before and after treatment samples based on measured protein levels and compared to single‐protein models. Proteins were ranked based on their individual performance and their gain of performance when combined with interaction partnersResultProteasome‐associated elements were overrepresented among proteins with top individual performance. Interaction models provided more details about related processes involving ubiquitination. Furthermore, various other immune, hormonal, and neuronal terms found to be enriched among the top interactors.ConclusionComplex biological functions can be mapped into pathways which are composed of and regulated by PPIs. Biological functionality can be encoded in concentrations of individual proteins, also in relative concentrations of interacting proteins. Our results indicate that the latter aspect may contain a significant amount of complementary information, not accessible with traditional approaches analyzing proteins as independent entities. Consequently, the development and the application of novel system biology tools on blood plasma proteomics will help to better understand patients’ responses to treatment in clinical trials.

Enhancing real-time nonintrusive occupancy estimation in buildings via knowledge fusion network

Real-time nonintrusive occupancy estimation can maximize the use of existing sensors to infer occupant information in buildings with the advantages of fewer privacy concerns and fewer extra device costs. Recently, many deep learning architectures have proven effective in estimating occupancy directly from raw sensor data. However, some handcrafted features manually extracted from statistical and temporal domains might convey additional information for occupancy estimation. In this study, a novel knowledge fusion network for nonintrusive occupancy estimation is proposed to integrate knowledge from two streams, i.e. automatic knowledge stream from a deep learning architecture and handcrafted knowledge stream from manual feature engineering. Moreover, four different fusion modules are investigated to optimize the design of the fusion network. To verify the effectiveness of the proposed network, experiments are conducted in a dataset from the ASHRAE Global Occupant Behavior Database, which is collected from an office space with records of indoor environment parameters, occupant-building interactions, and contextual information. The results demonstrate the superiority of the proposed fusion network, which outperforms five representative algorithms. Furthermore, the ablation study underscores the benefits of knowledge fusion and occupant-building interaction information, showing that the proposed fusion network can enhance the occupancy estimation accuracy by 3.47 % to 9.24 %.

Predicting drug-target binding affinity with cross-scale graph contrastive learning.

Identifying the binding affinity between a drug and its target is essential in drug discovery and repurposing. Numerous computational approaches have been proposed for understanding these interactions. However, most existing methods only utilize either the molecular structure information of drugs and targets or the interaction information of drug-target bipartite networks. They may fail to combine the molecule-scale and network-scale features to obtain high-quality representations. In this study, we propose CSCo-DTA, a novel cross-scale graph contrastive learning approach for drug-target binding affinity prediction. The proposed model combines features learned from the molecular scale and the network scale to capture information from both local and global perspectives. We conducted experiments on two benchmark datasets, and the proposed model outperformed existing state-of-art methods. The ablation experiment demonstrated the significance and efficacy of multi-scale features and cross-scale contrastive learning modules in improving the prediction performance. Moreover, we applied the CSCo-DTA to predict the novel potential targets for Erlotinib and validated the predicted targets with the molecular docking analysis.