Information Network Research Articles

Impact of Sensitive Circulating Tumor DNA Monitoring on CT Scan Intervals During Postoperative Colorectal Cancer Surveillance

Objective: This study investigated whether digital polymerase chain reaction (dPCR)-based circulating tumor DNA (ctDNA) monitoring can allow longer intervals between computed tomography (CT) scans during postoperative surveillance of colorectal cancer (CRC). Background: Practical guidelines still recommend intensive postoperative surveillance of CRC using periodical CT scans and serum carcinoembryonic antigen testing. Methods: The longitudinal dynamics of ctDNA for 52 patients with CRC as measured by dPCR using probes targeting 87 individual tumor-specific mutations (1–5 per patient) were compared with results from conventional (ie, clinical) surveillance using serum tumor markers and CT. Results: A total of 382 CT procedures were carried out for the patient cohort (3.3/year per patient) and the median lead time from ctDNA relapse to clinical relapse was 182 days (range, 0–376 days). If the CT interval was annual, potential delays in the detection of clinical relapse would have occurred for 7 of the 10 patients who experienced clinical relapse (9 of 13 events), with a median delay of 164 days (range, 0–267 days). If annual CT surveillance was performed together with ctDNA monitoring, 218 (57.1%) CTs would not have been needed to detect the first clinical relapse. In addition, the ctDNA monitoring would have provided a lead time of 339 days for detection of clinical relapse (range, 42–533 days). Conclusions: Our findings suggest that the ctDNA monitoring as part of postoperative surveillance and clinical relapse detection for patients with CRC could allow the CT interval to be lengthened. Trial Registration: This trial was registered with University Hospital Medical Information Network Clinical Trial Registry (UMIN000045114).

Price response to consumer attention during major animal epidemics in developing markets: perspectives from the influence of information and emotion contagion

Major animal epidemics can significantly disrupt the pork market. Understanding how consumer network attention, triggered by these epidemics, impacts pork price fluctuations is of great significance for maintaining market stability and ensuring food security. This study focuses on exploring this complex relationship, with a particular emphasis on the roles of information dissemination and emotion transmission. Taking African Swine Fever (ASF) as a case study, monthly provincial panel data from June 2021 to November 2022 were collected. Web scraping techniques and social network analysis were employed. Weibo user repost and emotion transmission networks were constructed, integrating social network structures into the analytical framework to comprehensively analyze the problem. The study reveals several important findings. Firstly, consumer network attention significantly intensifies pork price volatility during major epidemics, and there is notable heterogeneity across different information environments. Secondly, both information dissemination and emotion transmission play moderating roles. Specifically, the media information index negatively moderates the relationship between consumer attention and pork price fluctuations. The betweenness centrality of the Weibo repost network also has a negative moderating effect, indicating that “opinion leaders” on social media can mitigate the impact of consumer attention on price fluctuations through selective information dissemination. In contrast, the closeness centrality of the emotion transmission network has a positive moderating effect, highlighting the amplifying effect of rapid emotion propagation on market reactions. This research highlights the economic implications of the relationship between consumer network attention, information dissemination, emotion transmission, and pork price fluctuations during food safety incidents. It emphasizes the importance of understanding these dynamics for maintaining food market stability.

ZFIN Updates to Support Zebrafish Environmental Exposure Data.

The Zebrafish Information Network (ZFIN, zfin.org) is the database resource for genetic, genomic, and phenotypic data from research using zebrafish, Danio rerio. ZFIN curates information about genetic perturbations, gene expression, phenotype, gene function, and human disease models from zebrafish research publications and makes these data available to researchers worldwide. Over the past 20 years, zebrafish have increasingly been used to investigate the effects of environmental exposures, becoming an ideal model to study toxicity, phenotypic outcomes, and gene-chemical interactions. Despite this, database resources supporting zebrafish toxicology and environmental exposure research are limited. To fill this gap, ZFIN has expanded functionality to incorporate and convey toxicology data better. ZFIN annotations for gene expression, phenotype, and human disease models include information about genotypes and experimental conditions used. One type of experimental condition the database captures is the application of chemicals to zebrafish. ZFIN annotates chemicals using the Chemical Entities of Biological Interest Ontology (ChEBI) along with the Zebrafish Experimental Conditions Ontology (ZECO) to denote route of exposure and other experimental conditions. These features allow researchers to search phenotypes and human disease models linked to chemicals more efficiently. Here we discuss how experimental conditions are displayed on ZFIN web pages, the data displayed on chemical term pages, and how to search and download data associated with chemical exposure experiments.

Synthesizing the impact of armed conflicts on food security, livelihoods and social dynamics in Amhara region, Ethiopia

BackgroundArmed conflicts cause food insecurity and famine by triggering the disruption and destruction of food systems. The conflict that erupted in northern Ethiopia lasted from 2020 to 2022 and affected millions of people. This resulted in severe damage to the food security, livelihoods and social dynamics of communities in the Amhara region.ObjectiveThis study synthesized the impact of the conflict in northern Ethiopia on food security, livelihoods and social dynamics livelihoods in there zones, South Wello, North Wello and North Shewa zones of the Amhara Region. The conflict management response and the scope of humanitarian assistance are also analyzed.MethodsThe study used a qualitative case study design. Primary data were collected through semi-structured interviews, focus group discussions and observation. Data were collected from local authorities, representatives of displaced people, government and non-government officials who responded to the crisis and from the community members who were directly affected by the conflict. The primary data were supplemented by a review of reports from credible secondary sources. The research used thematic and narrative analysis techniques to synthesize the impact of armed conflicts on food security, livelihoods, and social dynamics in Amhara region, Ethiopia.ResultsOver 5.5 million people found to be affected by acute food insecurity and over half a million people were internally displaced due to the Northern conflict. The monetary value of the damage caused by the armed conflict is estimated at $500 million, due to the damage to basic services and the loss of 14 million quintals of agricultural production. The conflict has disrupted food systems by damaging agricultural infrastructure, causing market distortions and causing food shortages. Livelihoods were threatened by the loss of jobs and the destruction of private and municipal property. The destruction of the basic infrastructure such as health, water and sanitation, agriculture and market institutions inevitably have lasting negative impacts on the overall productivity of the community. Communities’ religious and traditional values, informal social networks ​​are affected by the impact of armed conflict, complicating communities’ reconstruction efforts.Conclusion and recommendationThe northern armed conflict has significant negative impacts on the food security, livelihoods and social dynamics of people in the Amhara region. Recognizing the multiple impacts of conflict livelihoods in the region is a critical step in developing effective responses. Government, non-governmental organizations, and community-based institutions should play a crucial role in building community resilience and support post-conflict reconstruction efforts. Future studies can capitalize on this qualitative finding and could design a large scale assessment using quantified data to further provide evidence on post-conflict reconstruction efforts.

“If It’s Not a Slap or a Punch Then It’s Not Violence”: Portuguese University Student’s Representations and Practices About Intimacies and Dating Violence

This study aimed to analyze characteristics, representations, and practices of intimacies and violence in Portuguese university students. Data collection involved 10 focus groups with a total of 35 university students aged 18–23. The findings reveal that students conceptualize healthy relationships based on respect, trust, and individuality, reflecting egalitarian ideals. However, participants acknowledge persistent gendered expectations and asymmetries. Contributing factors to dating violence (DV) include gender stereotyping, social media, substance use, and specific social settings. While social media can foster connection, it also enables control and online abuse. Substance use, although not a direct cause, may exacerbate pre-existing tensions. DV predominantly occurs in private settings, though emotional abuse also surfaces in public spaces like social gatherings. Participants condemned DV but noted the societal normalization of less visible forms of violence. They highlighted the role of peers as primary support networks for victims but also pointed to passive attitudes when abuse involves acquaintances. These insights underscore the need for targeted interventions addressing social norms and enhancing informal support networks in university settings. This study enriches the limited qualitative research on DV among Portuguese youth and emphasizes the importance of aligning prevention strategies with the nuanced realities of young people’s relationships.

A skeleton extraction method for large-scale spatial interaction networks considering spatial distribution characteristics

Constructing spatial interaction (SI) networks from massive geographic flow data is an important method for understanding the correlation patterns and trends among regions. However, large-scale SI networks usually cause complex edge interleaving and a sharp increase in noise information. A network skeleton can represent the key structure of the corresponding complex network with generalized network data. However, the existing extraction methods only consider the topological information of networks and ignore spatial information, which may cause difficulty maintaining the spatial distribution characteristics and the exclusion of key spatial information. To address the issue, we have developed a skeleton extraction method for large-scale SI networks. First, the network nodes are divided into different clusters via adaptive spatial clustering. Second, an entropy weight method is used to define a comprehensive evaluation indicator of the network node importance, and the nodes in each cluster are sorted and selected according to the comprehensive evaluation indicator. Finally, the skeleton structure of the SI network is reconstructed according to the selected nodes and their adjacent edges. Compared with the existing methods for extracting network skeletons, the proposed method can maintain the spatial distribution characteristics of SI networks while preserving key topological features.

Effects of home-based virtual reality upper extremity rehabilitation in persons with chronic stroke: a randomized controlled trial

BackgroundUpper extremity rehabilitation in persons with stroke should be dose-dependent and task-oriented. Virtual reality (VR) has the potential to be used safely and effectively in home-based rehabilitation. This study aimed to investigate the effects of home-based virtual reality upper extremity rehabilitation in persons with chronic stroke.MethodsThis was a single-blind, randomized, controlled trial conducted at two centers. The subjects were 14 outpatients with chronic stroke more than 6 months after the onset of the stroke. The participants were randomly divided into two groups. The intervention group (n = 7) performed a home rehabilitation program for the paretic hand (30 min/day, five days/week) using a VR device (RAPAEL Smart Glove™; NEOFECT Co., Yung-in, Korea) for four weeks. The control group (n = 7) participated in a conventional home rehabilitation program at the same frequency. All participants received outpatient occupational therapy once a week during the study period. The outcome measures included the Fugl-Meyer Assessment of upper extremity motor function (FMA-UE), Motor Activity Log-14 (MAL), Jebsen-Taylor Hand Function Test (JTT), and Box and Block Test (BBT) scores.ResultsAll 14 participants completed the study. Compared to the control group, the intervention group showed more significant improvements in FMA-UE (p = 0.027), MAL (p = 0.014), JTT (p = 0.002), and BBT (p = 0.014). No adverse events were observed during or after the intervention.ConclusionCompared to a conventional home program, combining a task-oriented virtual reality home program and outpatient occupational therapy might lead to greater improvements in upper extremity function and the frequency of use of the paretic hand.Trial registration: This study was registered in the University Hospital Medical Information Network (UMIN) Clinical Trial Registry in Japan (Unique Identifier: UMIN000038469) on November 1, 2019; https://center6.umin.ac.jp/cgi-open-bin/ctr/ctr_view.cgi?recptno=R000043836.

Effectiveness of Cognitive Behavioral Therapy Provided Through a Web Application for Subthreshold Depression, Subthreshold Insomnia, and Subthreshold Panic: Open-Labeled 6-Arm Randomized Clinical Trial Pilot Study.

A common definition of "subthreshold" is that the diagnostic threshold is not met but the individuals are not asymptomatic. Some symptoms are present, causing significant difficulty in functioning and negatively impacting quality of life. Despite the attention given to subthreshold symptoms and the interventions for subthreshold symptoms being efficient in preventing the transition to psychiatric disease in primary care, reports on specific interventions are insufficient. This study aimed to verify the effectiveness of internet-delivered cognitive behavioral therapy (ICBT) for subthreshold depression (SD), subthreshold insomnia (SI), and subthreshold panic (SP). Additionally, this study aimed to explore the minimally important change (MIC) of each subthreshold group's effectiveness outcome. Participants aged 18-70 years from internet research monitors were categorized into SD, SI, and SP groups based on screening assessment. They were randomly assigned to intervention or control groups within each subthreshold symptom. The intervention groups worked on 4 weeks of nonguided ICBT ("Mentre"), while the control groups worked on a sham app. The primary outcome was the score change from screening (T1) to 4-week follow-up (T4) using the Center for Epidemiologic Studies Depression Scale (CESD) in the SD group, the Pittsburgh Sleep Quality Index (PSQI) in the SI group, and the Panic and Agoraphobia Scale (PAS) in the SP group. Secondary outcomes were score changes in the Generalized Anxiety Disorder-7 (GAD-7) scale, the Patient Health Questionnaire 9 (PHQ-9), the CESD, the PSQI, and the PAS, except the primary outcome in each group. Secondary outcomes were analyzed using complete-case analysis and repeated-measures ANOVA. Additionally, the MIC in the primary endpoint for each group was also calculated as an exploratory outcome. The SD, SP, and SI groups contained 846, 597, and 1106 participants, respectively. In the SD group, the difference in the CESD score change from baseline to follow-up between the intervention and control groups was significant (difference=0.52, 95% CI 1.29-4.66, P<.001). In the SI group, the difference in the PSQI score change was also significant (difference=0.53, 95% CI 0.11-0.94, P=.01). However, in the SP group, the difference in the PAS score change was not significant (difference=0.07, 95% CI -2.00 to 2.15, P=.94). Our ICBT program Mentre contributes to the improvement of SI and SD. This suggests that nonguided ICBT may be effective in preventing SI and SD from progressing to the full threshold. However, appropriate definitions of subthreshold symptoms are necessary. In particular, it is difficult to define SP, and further research that considers the specific factors of each subthreshold symptom is necessary to accumulate evidence. University Hospital Medical Information Network (UMIN) UMIN000051280; https://tinyurl.com/2wyahhe3.

Informal Support Networks of Parents Involved With the Child Welfare System: Needs and Mental Health Symptoms.

Studies show the protective power of informal support networks for parents, however, most research in this area is not specific to the child welfare context. With a sample of parents with children involved with the U.S. child welfare system due to substantiated child maltreatment (N = 118), latent profile analyses revealed three distinct profiles of parents' informal support networks including perceived support, received support, and network demands. The profiles were associated with differences in sociodemographic risks, ACEs, and symptoms of depression and stress. Parents with four or more ACEs were more likely to be classified in the very little support profile versus the profile of parents with some support. Parents who had high levels of informal network support and a manageable level of network demands had lower levels of depression and stress compared to parents with very little network support. Implications for research and practice are discussed.

Locating the Source of Information in Social Networks using Critical Nodes

Locating the information source within social networks is crucial to understand information propagation. The source can be detected based on specific nodes known as observation nodes, and identifying them is a critical challenge that can significantly affect the accuracy of identification. To address this issue, this study proposes a novel source detection approach based on the Susceptible-Infected (SI) model and the Critical Node Problem (CNP). CNP involves identifying a subset of nodes within a graph whose removal results in the maximum reduction of a given connectivity metric, thereby isolating significant areas within the graph. A heuristic algorithm was developed, grounded in the maximal independent set for general graphs to solve the CNP, allowing the identification of the most crucial observation nodes that enhance the accuracy and using the data recorded from them to estimate the localization of the source. Experimental evaluations on various real-world networks showed that the proposed approach achieved a source detection accuracy of up to 89%, outperforming existing methods. These results demonstrate the robustness of the proposed approach, highlighting its potential to significantly improve accuracy in network-based source localization tasks across multiple applications.

Help-seeking behaviors among survivors of intimate partner violence during pregnancy in 54 low- and middle-income countries: evidence from Demographic and Health Survey data

BackgroundIntimate partner violence (IPV) during pregnancy is a global issue of public health importance. Due to the low rates of help-seeking in response to IPV, research on survivors’ help-seeking behaviors is scarce, particularly within low- and middle-income countries (LMICs). The present study provides a cross-national evidence of informal and formal help-seeking patterns among pregnant women experiencing IPV.MethodsThis study made use of population-based data from the Demographic and Health Surveys (DHS) Program from 54 LMICs, collected between 2005 and 2020 (N = 359,027). Applying bivariate and multivariable analyses, the present study examined IPV survivors’ help-seeking distributions and associations with individual, partner, family, and community factors.ResultsOnly half of survivors sought help in response to IPV during pregnancy with wide regional and national variations, mainly from informal support networks (including family, neighbors, and friends), and rarely from formal institutions (including legal, socio-cultural, and medical services). Evidence shows that help-seeking behaviors were associated with IPV survivors’ age and educational attainment, survivors’ employment status and earnings compared to their partners, survivors’ consumption of mass media, intimate partner’s age and education, spouse’s alcohol consumption and controlling behaviors, survivors’ wealth index, place of residence, and health-seeking barriers, among others.ConclusionPractitioners are encouraged to consider the study’s outcomes when designing interventions and support for survivors seeking help in response to IPV during pregnancy. Strong advocacy and action are needed, including fostering survivors’ educational attainment, diminishing pregnant women’s social and financial dependencies on their intimate partners, promoting pre- and peri-natal health care, informing survivors about help, and increasing gender equality by engaging women and men equally within the whole community.

A Social Network Lens to Community Health Worker Influence and Impact.

Community Health Workers (CHWs) are members of healthcare teams that are integrated in, and often share language, beliefs, and lived experiences with their communities. They use their formal and informal social networks to promote healthy behavior, to connect community members to resources, and to build more resilient community networks. We propose a framework to conceptualize CHW interventions aiming to operationalize and optimize CHW social relations and networks. CHW-mediated network interventions can focus on the dissemination and diffusion of health messages, using the channels of trust and formal and informal relations, as well as, engaging communities to enhance the cascade of spreading/diffusion. Network interventions can also focus on network-building and community dialog, relying on the role of CHWs in bringing the community members together in facilitating conversation, promoting social justice and inclusion, and mobilizing the community in collective action. In addition, the network interventions can aim for boundary-spanning and bridging activities, to facilitate the community's access to health services and external resources, as well as bringing the community voice to health systems to influence priorities and policies. Similar to any other complex interventions, CHW network interventions should be fine-tuned and adapted to local and community needs, capacities, and network structures, and actively involve community members in the conceptualization, delivery, and evaluation.

Do Survivor and Supporter Perceptions of Supporter Helpfulness and Social Reactions Affect Survivor PTSD Symptoms?

Prior studies indicate that sexual assault survivor posttraumatic stress disorder (PTSD) is influenced by the social reactions (e.g., belief and blame) and perceived helpfulness (e.g., empathy, effectiveness) of informal support providers (SPs; i.e., family, friends, and partners). This study examined how SPs' social reactions and perceived helpfulness impacted female sexual assault survivor PTSD among an ethnically diverse sample of 432 pairs of sexual assault survivors and their SPs. Path models were run separately for survivors and SPs. In the survivor model, SP's perceived ineffectiveness was significantly related to more unsupportive acknowledgment and turning against reactions, and perceived empathy was significantly related to more unsupportive acknowledgment reactions. Both turning against and unsupportive acknowledgment reactions were related to greater survivor PTSD. In the SP model, greater SP confusion about how to help the survivor was related to marginally less SP unsupportive acknowledgment and more turning against social reactions made to survivors. Greater SP empathy toward the survivor was also related to significantly less SP unsupportive acknowledgment and turning against social reactions. Finally, SP turning against social reactions was related to marginally more survivor PTSD symptoms. Results suggest that longitudinal research is needed to explore further how helping perceptions relate to survivor PTSD from both survivor and SP perspectives to inform informal support network interventions. Interventions to increase empathy toward sexual assault survivors and teach SPs how to avoid responding with negative social reactions are needed.

AGHINT: Attribute-guided representation learning on heterogeneous information networks with transformer

AGHINT: Attribute-guided representation learning on heterogeneous information networks with transformer

Enhanced detection of obfuscated HTTPS tunnel traffic using heterogeneous information network

Enhanced detection of obfuscated HTTPS tunnel traffic using heterogeneous information network

Kidney Biopsy-Proven Diabetic and Non-Diabetic Kidney Diseases and Outcomes in Patients With Type 2 Diabetes Receiving Dialysis: The REIN Registry.

Chronic kidney disease (CKD) in patients with diabetes does not always equate to diabetic kidney disease (DKD). This study aims to delineate and compare the clinical characteristics, survival rates, and access to kidney transplantation among patients with type 2 diabetes commencing dialysis, who were classified by kidney biopsy as having either DKD or non-diabetic kidney disease (non-DKD). We used the comprehensive French Renal Epidemiology and Information Network registry to analyze baseline clinical characteristics at dialysis inception and outcomes defined as death and access to kidney transplantation. We employed a multivariate Cox proportional hazards model and the Fine-Gray competing risk model to assess the probabilities of mortality and transplantation. Adults in the Renal Epidemiology and Information Network registry in France with a diagnosis of type 2 diabetes who initiated kidney replacement therapy from January 2009 to December 2015 and had a previous native kidney biopsy. We analyzed data from 2,869 patients with diabetes, 45% of whom had a biopsy-confirmed diagnosis of DKD. Among these patients, half presented additional histopathological findings indicative of nephroangiosclerosis and focal segmental glomerulosclerosis. The clinical profiles of patients with DKD and non-DKD were largely comparable. There were no significant differences in dialysis survival rates or kidney transplantation access between the groups, even after adjusting for confounding variables and considering competing risks. At the 6-year mark, the mortality rate was 60.3% (95% CI: 55.5-64.5) for the DKD group and 60.3% (95%CI: 55.9-64.3) for the non-DKD group. Multivariable Cox analysis revealed no significant difference in mortality risk between the DKD and non-DKD groups. The study limitations include potential residual confounders, lack of predialysis data, kidney biopsies possibly outdated, nonrandom biopsy indications, and survival bias because of analysis at dialysis inception. In patients with diabetes initiating dialysis, clinical characteristics and outcomes following dialysis initiation were similar in biopsy-proven DKD versus non-DKD. Our results suggest that the diabetic milieu has a more significant impact on outcomes in patients with diabetes treated with dialysis than the underlying pathological kidney diagnosis.

An integrated neighborhood and scale information network for open-pit mine change detection in high-resolution remote sensing images

An integrated neighborhood and scale information network for open-pit mine change detection in high-resolution remote sensing images

BARE: Balance representation for imbalance multi-class node classification on heterogeneous information networks

BARE: Balance representation for imbalance multi-class node classification on heterogeneous information networks

Multi-Agent Deep Reinforcement Learning with Graph Attention Network for Traffic Signal Control in Multiple-Intersection Urban Areas

Deep reinforcement learning has seen significant progress in traffic signal control. However, existing research still lacks the ability to effectively capture the correlation of road network information and the perception capability of traffic signal states. To address this gap, we propose a multi-intersection traffic signal control method that integrates a graph attention network, named the graph attention network-deep deterministic polcy gradient (GAT-DDPG) algorithm. This algorithm incorporates the restart random walk into the attention mechanism, exploring graph information through global random walks, reducing reliance on local nodes, and enhancing the model’s comprehensive understanding of graph structure features, thereby improving the modeling capability of traffic network structures. Moreover, the algorithm can automatically identify and extract key features from the complex data of the traffic network without manual intervention, adapting to different traffic network topologies, and can update and adjust the traffic signal control system in real-time to accommodate actual traffic flow and congestion situations. Experimental results indicate that the GAT-DDPG algorithm reduces average vehicle travel time significantly across three real road networks (Hangzhou and Jinan in China, and New York, U.S.) and two synthetic road network datasets. Additionally, it demonstrates optimal convergence speed and performance in these real datasets, attributed to its capability to capture global information and deeply comprehend the intricate structures of traffic networks. The research proves that this model has significant advantages in the field of traffic signal control, improving the operational efficiency of urban area intersections. Future work will incorporate additional road environment factors to better adapt to complex urban traffic.

Truncated Arctangent Rank Minimization and Double-Strategy Neighborhood Constraint Graph Inference for Drug-Disease Association Prediction.

Accurately identifying new therapeutic uses for drugs is essential to advancing pharmaceutical research and development. Graph inference techniques have shown great promise in predicting drug-disease associations, offering both high convergence accuracy and efficiency. However, most existing methods fail to sufficiently address the issue of numerous missing information in drug-disease association networks. Moreover, existing methods are often constrained by local or single-directional reasoning. To overcome these limitations, we propose a novel approach, truncated arctangent rank minimization and double-strategy neighborhood constraint graph inference (TARMDNGI), for drug-disease association prediction. First, we calculate Gaussian kernel and Laplace kernel similarities for both drugs and diseases, which are then integrated using nonlinear fusion techniques. We introduce a new matrix completion technique, referred to as TARM. TARM takes the adjacency matrix of drug-disease heterogeneous networks as the target matrix and enhances the robustness and formability of the edges of DDA networks by truncated arctangent rank minimization. Additionally, we propose a double-strategy neighborhood constrained graph inference method to predict drug-disease associations. This technique focuses on the neighboring nodes of drugs and diseases, filtering out potential noise from more distant nodes. Furthermore, the DNGI method employs both top-down and bottom-up strategies to infer associations using the entire drug-disease heterogeneous network. The synergy of the dual strategies can enhance the comprehensive processing of complex structures and cross-domain associations in heterogeneous graphs, ensuring that the rich information in the network is fully utilized. Experimental results consistently demonstrate that TARMDNGI outperforms state-of-the-art models across two drug-disease datasets, one lncRNA-disease dataset, and one microbe-disease dataset.