SEARCH · PubMed Health

Results for “multimodal learning”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3Linked to original sources

PubMed2026

AI-driven diagnostic and prognostic models for metabolic dysfunction-associated steatotic liver disease: insights from clinical, imaging, and multi-omics studies-a scoping review.

Metabolic dysfunction-associated steatotic liver disease (MASLD), formerly known as non-alcoholic fatty liver disease (NAFLD), is the most common chronic liver disease around the world, affecting 33.6% of the adult population (95% CI: 28.1%-39.5%; I 2 = 99.9%), or roughly one in three. The extent of the liver damage is variable, from simple steatosis to metabolic dysfunction-associated steatohepatitis (MASH, formerly NASH), cirrhosis and hepatocellular carcinoma (HCC). Early diagnosis is essential to prevent serious liver damage. Traditional diagnostic techniques such as liver biopsy, imaging, and biomarker testing are all invasive, costly, reduced sensitive to early-stage disease, and they also have variability among observers. Modern diagnostic and prognostic approaches based on the principles of Artificial Intelligence (AI) and specifically on machine learning (ML) and deep learning (DL) have enabled multimodal approaches integrating clinical, imaging and molecular data. This scoping review conducted per PRISMA-ScR guidelines, synthesizes findings from 73 studies (search window 2020-2026) across three dimensions: clinical data driven models, imaging-based classifiers (ultrasound, CT and MRI), and multi-omics (genomics, transcriptomics and proteomics) techniques. Moreover, emergence of models such as U-Net and LiverNet 2.x, classification models like DeepLiverNet and BiLSTM models, as well as transformer frameworks and the identification of biomarkers models are also described. This study also investigates challenges such as data heterogeneity, data interpretability, fairness and real-world clinical application. Finally, important areas of research opportunities and future directions are highlighted to present a developing clinically applicable, explainable and ethical AI solutions to manage MASLD.

MASLD↗

PubMed2026

GBFN: A gated bimodal fusion network leveraging foundation model embeddings for cancer drug sensitivity prediction.

Despite recent progress in deep learning for cancer drug sensitivity prediction, many existing models still rely on task-specific representation learning or relatively simple multimodal fusion, which may limit their ability to capture complex drug-cell interactions. To address this issue, we developed GBFN, a gated bimodal fusion network for continuous IC50 prediction that integrates pretrained drug and cell-line representations. Specifically, drug embeddings were obtained from SMI-TED, whereas cell-line embeddings were derived from transcriptomic profiles using BulkFormer. These two modalities were then combined through a dimension-wise gated fusion module and used to predict IC50 values in matched drug-cell line pairs. On the CCLE-based benchmark, GBFN outperformed representative neural baselines, including GraphDRP, TGSA, and TransEDRP, and achieved the best overall performance, with an R² of 0.8714 and an RMSE of 0.8938. Moreover, ablation analysis showed that the model using drug features and cell-line expression data with gated fusion performed better than the corresponding model using direct concatenation, indicating that the improvement was associated with the fusion strategy rather than with the input modalities alone. In addition, cell-line expression data were more informative than mutation data in the present setting, and adding mutation data to the model using drug features and expression data did not further improve performance. Across major cancer types, GBFN maintained generally high cell-line-level predictive performance, and perturbation-based attribution identified biologically relevant transcriptomic programs in selected drug-cell line settings. Together, these findings support GBFN as a compact and effective framework for continuous drug response prediction.

Humans↗

PubMed2002

Management of chronic prostatitis-chronic pelvic pain syndrome.

Although the neurobiologic basis of CPPSs in men remains unclear, therapeutic interventions should continue to be improved. Invasive or destructive modalities should be avoided when possible. Electrical neuromodulation techniques seem to be a promising, among other multimodal approaches. Physicians must learn from patients in attempt to relieve symptoms.

Biofeedback, Psychology↗

PubMed1994

Odorant-induced oscillations in the mushroom bodies of the locust.

Kenyon cells are the intrinsic interneurons of the mushroom bodies in the insect brain, a center for olfactory and multimodal processing and associative learning. These neurons are small (3-8 microns soma diameter) and numerous (340,000 and 400,000 in the bee and cockroach brains, respectively). In Drosophila, Kenyon cells are the dominant site of expression of the dunce, DC0, and rutabaga gene products, enzymes in the cAMP cascade whose absence leads to specific defects in olfactory learning. In honeybees, the volume of the mushroom body neurophils may depend on the age or social status of the individual. Although the anatomy of these neurons has been known for nearly a century, their physiological properties and the principles of information processing in the circuits that they form are totally unknown. This article provides a first such characterization. The activity of Kenyon cells was recorded in vivo from locust brains with intracellular and local field potential electrodes during olfactory processing. Kenyon cells had a high input impedance (approximately 1 G omega at the soma). They produced action potentials upon depolarization, and consistently showed spike adaptation during long depolarizing current pulses. They generally displayed a low resting level of spike activity in the absence of sensory stimulation, despite a large background of spontaneous synaptic activity, and showed no intrinsic bursting behavior. Presentation of an airborne odor, but not air alone, to an antenna evoked spatially coherent field potential oscillations in the ipsilateral mushroom body, with a frequency of approximately 20 Hz. The frequency of these oscillations was independent of the nature of the odorant. Short bouts of oscillations sometimes occurred spontaneously, that is, in the absence of odorant stimulation. Autocorrelograms of the local field potentials in the absence of olfactory stimulation revealed small peaks at +/- 50 msec, suggesting an intrinsic tendency of the mushroom body networks to oscillate at 20 Hz. Such oscillatory behavior could not be seen from local field potential recordings in the antennal lobes, and may thus be generated in the mushroom body, or via feedback interactions with downstream neurons in the protocerebrum. During the odor-induced oscillations, the membrane potential of Kenyon cells oscillated around the resting level, under the influence of excitatory inputs phase-locked to the field activity. Each phasic wave of depolarization in a Kenyon cell could be amplified by intrinsic excitable properties of the dendritic membrane, and sometimes led to one action potential, whose timing was phase-locked to the population oscillations.(ABSTRACT TRUNCATED AT 400 WORDS)

Animals↗

PubMedDate not supplied

A layered architecture for computer-based simulation supporting skills learning: an X-ray imaging paradigm.

Simulation is characterized by strong learning potential, providing the basis for a new category of systems, the simulation-based learning systems. To strengthen the learning potential of these systems, models are needed not only of the actual system being imitated, but also of the operational expertise required to carry out manipulations of the simulated system, inherently linked to learning. In this paper, an architecture is reported aimed at supporting the organization of multimodal simulation resources to induce skills learning. This architecture is based on distinct layers, allowing independent representation of learning and simulation components. Its applicability has been demonstrated by means of a paradigm, including simulation of X-ray imaging procedure, as well as authoring of learning scenarios pertaining to such procedures.

Algorithms↗

PubMed2026

GICPIdb: an archival repository of multimodal data focusing on pathological images for gastrointestinal cancers.

INTRODUCTION: Deep learning (DL) shows great potential for predicting biomarkers from routine histopathological slides of gastrointestinal (GI) cancers. Yet most existing models are validated on limited patient cohorts, while pathological image annotation and molecular marker standardization demand substantial professional expertise. To address these gaps, we constructed the Gastrointestinal Cancer Pathological Image Archive (GICPIdb, gicpidb.shubuzuo.top), a dedicated database and web platform covering seven major GI cancer types. METHODS: High-quality hematoxylin and eosin (H&E)-stained whole-slide images were collected from multiple sources and uniformly processed. Image annotations were performed by board-certified pathologists following standardized protocols. GICPIdb offers five interactive web modules for data uploading, quality control, feature extraction, online annotation and AI-based prediction. Its intuitive interface supports data browsing, retrieval, visualization and downloading. RESULTS: The database houses 2,863 pathologist-annotated, uniformly processed, high-quality H&E stained images collected from 2,655 patients. Of these, 1,699 patients were sourced from The Cancer Genome Atlas (TCGA), 182 from the Clinical Proteomic Tumor Analysis Consortium (CPTAC), and 424 from China-Japan Friendship Hospital and 350 from Chifeng Municipal Hospital in Inner Mongolia, China. It also integrates data on over 50 key molecular markers (e.g., MSI, TMB) and prognostic labels related to survival, recurrence and metastasis. DISCUSSION: GICPIdb aims to promote the development of DL-driven AI tools for cancer research and clinical translation. The multi-institutional data collection and standardized annotation pipeline are expected to enhance the generalizability and reproducibility of AI-based prediction models across diverse patient populations.

deep learning↗

PubMed2005

Learning preferences of caregivers of asthmatic children.

BACKGROUND: People learn in different ways: visually, aurally, by reading/writing, and kinesthetically. In our clinic, we use color-coded Asthma Action Cards to educate our patients and their caregivers on asthma management. Our teaching is largely aural based, with the cards providing reading and visual stimulation and hands-on practice with devices offering kinesthetic stimulation. OBJECTIVE: We sought to determine the learning styles of the caregivers of our asthmatic children. METHODS: Caregivers in our Asthma/Allergy Clinic completed the Visual-Aural-Read/Write-Kinesthetic (VARK) questionnaire anonymously, and the responses were evaluated on the basis of previously validated scoring instructions. RESULTS: Analysis of 98 respondents showed that 42% had a single learning modality preference, and the remaining 58% were multimodal learners. Of those who reported a single mode of learning, 61% preferred kinesthetic, 27% preferred reading/writing, and less than 1% each preferred aural or visual stimuli. Of all 98 caregivers, 82% included kinesthetic as a learning preference, 59% included read/write, 50% included aural, and 41% included visual. CONCLUSION: The majority of caregivers preferred the kinesthetic learning method, whether as a single learning preference or in combination with other approaches. Incorporating kinesthetic methods of learning, such as role plays and problem-solving case scenarios, into standardized asthma education curricula may be beneficial to patients and families in terms of understanding and using their regimen.

Asthma↗

PubMed2026

Multimodal features and prognostic risk assessment in locally advanced gastric cancer patients following neoadjuvant therapy based on machine learning algorithms: a multicenter study.

BACKGROUND: Neoadjuvant therapy (NAT) is recommended for locally advanced gastric cancer (LAGC), but some patients respond poorly. We aimed to construct a multimodal model integrating CT images, transcriptomic sequencing, and clinicopathological data to assess prognosis in LAGC patients receiving NAT. MATERIALS AND METHODS: This multicenter study included 505 LAGC patients who underwent NAT. Radiomic features were extracted from preoperative CT images of 505 patients. RNA-seq was performed on 277 post-NAT specimens, with additional data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases (n = 804). Patients were divided into training (168 cases), internal validation (72 cases), and external validation cohorts. Machine learning algorithms identified key radiomic, molecular, and clinical features associated with NAT response, which were then integrated into a multimodal model to predict overall survival (OS) and disease-free survival (DFS). RESULTS: Six radiomic and three molecular features significantly associated with NAT response were selected. Radiomic risk (hazard ratio [HR]: 4.0, P < 0.001) and molecular risk (HR: 7.1, P < 0.001) were independent prognostic factors. By integrating radiomic risk, molecular risk, and clinical characteristics, a multimodal model (MuMo) was constructed.The C-index results (OS, C-index = 0.855; DFS, C-index = 0.786) demonstrated that MuMo outperformed the single-modality models and ypTNM staging.Mechanistic analysis suggested that the efficacy of neoadjuvant therapy was significantly enriched in immune-inflammatory pathways. CONCLUSIONS: MuMo can effectively predict postoperative survival risk in LAGC patients receiving NAT, serving as a powerful tool for optimizing prognostic assessment.

Humans↗

PubMed2026

Machine learning for population-level risk prediction of future cholangiocarcinoma.

BACKGROUND: The poor prognosis of cholangiocarcinoma (CCA) is largely driven by rapid, asymptomatic disease progression, which usually results in a late diagnosis in the absence of established screening strategies. An early, cost-effective, and universally applicable risk assessment strategy would therefore be valuable. METHODS: We developed machine learning (ML) models on prospective, multimodal data from 487,495 UK Biobank (UKB) participants, of whom 649 developed CCA during follow-up. Data from England (80%) were utilised for ML development via five-fold cross-validation, and then all models were tested on withheld data from Scotland, Wales, and Newcastle (20%). Iterative ablation studies reduced inputs from >150 features across demographic data, lifestyle, health records, blood parameters, genomics, and metabolomics to models built on five and ten routinely available clinical parameters. These were externally validated in the Penn Medicine Biobank (PMBB; n = 2638; 28 CCA), All of Us Research Program (AOU; n = 330,433; 362 CCA), Japan Medical Data Centre Claims Database (JMDC; n = 8,425,522; 723 CCA) and TriNetX (n = 728,886; 1592 CCA). FINDINGS: We show that ML models integrating biliary-disease associated health records and Gamma glutamyltransferase can stratify risk of future CCA. Evaluation on the UKB test set as well as three independent cohorts revealed robust performance and generalisability across ethnicities. We achieved AUROCs of 0.71 [95% CI: 0.703-0.711], 0.77 [95% CI: 0.764-0.778 ], 0.796 [95% CI: 0.795-0.798] and 0.8 [95% CI: 0.794-0.805] for UKB, PMBB, AOU, and JMDC respectively, with respective AUPRCs of 0.014 [95% CI: 0.009-0.018], 0.042 [95% CI: 0.037-0.048], 0.038 [95% CI: 0.033-0.042] and 0.001 [95% CI: 0.001-0.001]. In AOU, application of the Youden J-optimised threshold yielded a number needed to screen of 79. Separate models for intra- and extrahepatic CCA did not improve performance. In line with the pathophysiology, performance declined for longer intervals between assessment and event. A group-level analysis in the TriNetX cohort revealed hazard ratios of up to 82.5 [95% CI: 26.4-257.96]. We provide extensive interpretability results and release all source codes used to develop the presented models. INTERPRETATION: We provide a comprehensive framework for early CCA risk stratification in the general population, identifying key predictors, and demonstrating the potential of data-driven models in personalised screening for hepatobiliary cancer. FUNDING: German Cancer Aid (grant #70115730), Junior Principal Investigator Fellowship programme of RWTH Aachen Excellence strategy.

Humans↗

PubMed2005

Clean intermittent catheterization in genitally sensate children: patient experience and health related quality of life.

PURPOSE: Clean intermittent catheterization (CIC) has proven successful for bladder emptying in children without genital sensation with success rates of 94% to 100% in select groups. A subset of the pediatric population requires CIC for bladder dysfunction, yet has normal genital sensation. This study was designed to assess our experience with CIC in genitally sensate children and the health related (HR) quality of life (QOL) for them and their families. MATERIALS AND METHODS: A multimodality questionnaire on the usage, learning curve and degree of difficulty of CIC was developed. The PedsQLtrade mark 4.0 Generic Core Scales, a reliable, validated HRQOL survey developed for pediatric patients with chronic disease, was used to assess patient and parental QOL. Children in our practice with normal genital sensation and bladder dysfunction requiring CIC were contacted to complete these questionnaires. When possible, QOL data were collected from the patient (self-report) and one or both of the parents (parent-proxy report). RESULTS: Data for 12 males and 8 females were collected, including 30 QOL questionnaires (12 self, 18 proxy). Of the patients 80% were able to learn CIC technique in 1 clinic visit with the help of a nurse. On a 10-point scale (1-not difficult, 10-very difficult) the mean degree of difficulty for learning the CIC technique was 3.25 for males and 4.00 for females (3.55 overall). On a 10-point scale (1-uncomfortable, 10-very comfortable) mean comfort level with the CIC technique was 9.12 for females and 9.46 for males (9.33 overall). On the PedsQL 100-point scale mean QOL was 83.24 for the self-report compared to 81.78 for the parent proxy report. The QOL for normal children has previously been shown as 83 (self) to 87.61 (proxy). CONCLUSIONS: CIC was an easy technique for most sensate children to learn in 1 visit and master in a short time. Overall comfort with the technique was excellent and few problems were encountered. Their HRQOL was comparable to that of normal children.

Adolescent↗

PubMed2006

Chronic obstructive pulmonary disease, risk factors, and outcome trials: comparisons with cardiovascular disease.

Chronic obstructive pulmonary disease (COPD) is a major health problem and now ranks fifth in terms of the global burden of disease. Although COPD is a disease that is characterized by progressive respiratory symptoms and functional decline, exacerbations pose the greatest risk for morbidity and early mortality, have a dramatic effect on quality of life, and are the most significant source of health care expenditure. To improve survival and reduce costs, it is critical to develop effective programs designed to reduce the frequency and severity of exacerbations for these patients. With limited health care resources, efficient and effective management of COPD ideally involves identifying and focusing efforts on individuals at particular risk. In the development of an appropriate multimodal strategy, lessons could be learned from the evolution of guidelines and management of cardiovascular disease, in particular heart failure, which has many parallels with COPD in terms of prevalence, prognosis, and impact on patient quality of life. There is a need for large prospective trials in COPD, based on hard clinical outcomes such as death, which, together with physician and patient education, will help to drive improvements in clinical management.

Cardiovascular Diseases↗

PubMedDate not supplied

Integrative properties of the Pe1 neuron, a unique mushroom body output neuron.

A mushroom body extrinsic neuron, the Pe1 neuron, connects the peduncle of the mushroom body (MB) with two areas of the protocerebrum in the honeybee brain, the lateral protocerebral lobe (LPL) and the ring neuropil around the alpha-lobe. Each side of the bee brain contains only one Pe1 neuron. Using a combination of intracellular recording and neuroanatomical techniques we analyzed its properties of integrative processing of the different sensory modalities. The Pe1 neuron responds to visual, mechanosensory, and olfactory stimuli. The responses are broadly tuned, consisting of a sustained increase of spike frequency to the onset and offset of light flashes, to horizontal and vertical movements of extended objects, to mechanical stimuli applied to the antennae or mouth parts, and to all olfactory stimuli tested (29 chemicals). These multisensory properties are reflected in its dendritic organization. Serial reconstructions of intracellularly stained Pe1 neurons using confocal microscopy reveal that the Pe1 neuron arborizes throughout all layers of MB peduncle with finger-like, vertically oriented dendrites. The peduncle of the MB is formed by the axons of Kenyon cells, whose dendritic inputs are organized in modality-specific subcompartments of the calyx region. The peduncular arborization indicates that the Pe1 neuron receives input from Kenyon cells of all calycal subcompartments. Because the Pe1 neuron changes its odor responses transiently as a consequence of olfactory learning, we hypothesize that the multimodal response properties might have a role in memory consolidation and help to establish contextual references in the long-term trace.

Adaptation, Physiological↗

PubMed2003

Abnormal variability and distribution of functional maps in autism: an FMRI study of visuomotor learning.

OBJECTIVE: Autism is a neurally based psychiatric disorder, but there is no consensus regarding the underlying neurofunctional abnormalities. Previous functional magnetic resonance imaging (fMRI) studies of simple movement suggested individually variable and scattered functional brain organization in autism. The authors examined whether such abnormalities generalize to multimodal processing (visually driven motor sequence learning). METHOD: Eight male autistic patients and eight comparison subjects matched with the patients on age, gender, and handedness were examined by using fMRI while they performed finger press movements prompted by visually presented repeating six-digit sequences. Hemodynamic responses to the six-digit sequences were statistically compared to responses to single-digit stimuli in one experiment and to regular six-digit sequences in another experiment. RESULTS: Both groups showed activations in bilateral premotor, superior parietal, and occipital cortices in both experiments. Task-by-group interactions showed that superior parietal activations were less pronounced in the autism group, whereas prefrontal cortex and more posterior parietal loci showed greater activation in the autism group than in the comparison group. The distances between Individual subjects' activation peaks and the groupwise peak were greater in the autism group than in the comparison group. CONCLUSIONS: The results support earlier findings of abnormal variability and scatter of functional maps in autism. They are consistent with evidence from other studies suggesting early-onset disturbances in the development of cerebello-thalamo-cortical pathways in autism.

Adolescent↗

PubMed2002

Prenatal experience and postnatal perceptual preferences: evidence for attentional-bias in bobwhite quail embryos (Colinus virginianus).

Previous studies have indicated that concurrent multimodal sensory stimulation can interfere with prenatal perceptual learning. This study further examined this issue by exposing 3 groups of bobwhite quail embryos (Colinus virginianus) to (a) no supplemental stimulation, (b) a bobwhite maternal call, or (c) a maternal call paired with a pulsating light in the period prior to hatching. Experiments differed in terms of the types of stimuli presented during postnatal preference tests. Embryos receiving no supplemental stimulation showed no preference between stimulus events in all testing conditions. Embryos receiving exposure to the unimodal maternal call preferred the familiar call over an unfamiliar call regardless of the presence or absence of pulsating light during testing. Embryos exposed to the call-light compound preferred the familiar call only when it was paired with the light during testing. These results suggest that concurrent multimodal stimulation does not interfere with prenatal perceptual learning by overwhelming the young organism's limited attentional capacities. Rather, multimodal stimulation biases what information is attended to during exposure and subsequent testing.

Animals↗

PubMed2006

Roles of the auditory cortex in discrimination learning by rats.

We investigated the roles of the auditory cortex in sound discrimination learning in Wistar rats. Absolute pitch or relative pitch can be used as discrimination cues in sound frequency discrimination. To clarify this, rats were trained to discriminate between rewarded (S+) and unrewarded (S-) test stimuli (S+ frequency>S- frequency). After learning was acquired by rats, performance was tested in a new test in which S+ frequency was constant but S+ frequency S- frequency but both frequencies were increased. If the discrimination cue of the first test was preserved in the new test, performance following change of testing procedures was expected to remain high. The measured performance suggested that rats used relative pitch in half octave discrimination (difference between S+ and S- frequencies, 0.5 octave), and absolute pitch in octave discrimination (difference between S+ and S- frequencies, 1.0 octave). Bilateral lesions in the auditory cortex had almost no effect on performance before procedure change. Furthermore, performance following procedure change was not affected by lesions in the auditory cortex when the discrimination cue was preserved. However, performance was impaired by lesions in the auditory cortex when a new discrimination cue was used following procedure change. Lesions in the auditory cortex also impaired multimodal discrimination between sound and sound plus light. The present findings suggest that the auditory cortex plays a role as a sensory interface of the higher cortices required for flexible learning and multimodal discrimination.

Acoustic Stimulation↗

PubMed2002

From view cells and place cells to cognitive map learning: processing stages of the hippocampal system.

The goal of this paper is to propose a model of the hippocampal system that reconciles the presence of neurons that look like "place cells" with the implication of the hippocampus (Hs) in other cognitive tasks (e.g., complex conditioning acquisition and memory tasks). In the proposed model, "place cells" or "view cells" are learned in the perirhinal and entorhinal cortex. The role of the Hs is not fundamentally dedicated to navigation or map building, the Hs is used to learn, store, and predict transitions between multimodal states. This transition prediction mechanism could be important for novelty detection but, above all, it is crucial to merge planning and sensory-motor functions in a single and coherent system. A neural architecture embedding this model has been successfully tested on an autonomous robot, during navigation and planning in an open environment.

Animals↗

PubMed2025

JASMINE: A powerful representation learning method for enhanced analysis of incomplete multi-omics data.

Integrative analysis of multi-omics data provides a more comprehensive and nuanced view of a subject's biological state. However, high-dimensionality and ubiquitous modality missingness present significant analytical challenges. Existing methods for incomplete multi-omics data are scarce, do not fully leverage both modality-specific and shared information, and produce task-biased representations. We propose JASMINE, a self-supervised representation learning method for incomplete multi-omics data that preserves both modality-specific and joint information and enhances sample similarity structure. JASMINE produces embeddings that achieve superior performance across multiple tasks for two different incomplete multi-omics datasets while requiring only a single round of training per dataset.

missing data↗

PubMed2025

Target and biomarker exploration portal for drug discovery.

MOTIVATION: The discovery of novel drug targets and precision biomarkers remains a major challenge in drug development, with traditional differential expression analysis often overlooking key regulatory proteins. Here, we present a novel, web-based bioinformatics tool, the Target and Biomarker Exploration Portal (TBEP), designed to accelerate the drug discovery process by integrating large-scale biomedical data with network analysis techniques. RESULTS: TBEP harnesses machine-learning approaches to mine and combine multimodal datasets, including human genetics, functional genomics, and protein-protein interaction networks, to decode causal disease mechanisms and uncover novel therapeutic targets and precision biomarkers for specific phenotypes. A unique feature of the tool is its ability to process large-scale data in real-time, facilitated by an efficient cloud-based architecture. Additionally, the tool incorporates an integrated large language model (LLM), which assists researchers in exploring and interpreting complex biological relationships within the generated networks and multi-omics data using natural language (English). By offering an intuitive, interactive interface, the LLM enhances the exploration of biological insights, making it easier for scientists to derive actionable conclusions. This powerful integration of network analysis, multi-omics data, and LLM provides a robust framework for accelerating the identification of novel drug targets. AVAILABILITY AND IMPLEMENTATION: The tool is publicly available at https://tbep.missouri.edu. The source code, documentation and installation instructions are available at GitHub repository: https://github.com/mizzoudbl/tbep.

Drug Discovery↗