Biology subjects

Bozdag, S.

Publications and source records attributed to Bozdag, S..

At least 19 recordsLinked to original sources

bioRxiv2026

From General-Purpose to Disease-Specific Features: Aligning LLM Embeddings on a Disease-Specific Biomedical Knowledge Graph for Drug Repurposing

Identifying new therapeutic uses for existing drugs is a major challenge in biomedicine, especially for complex neurodegenerative conditions such as Alzheimer disease and related dementias (ADRD), where treatment options remain limited and relevant data are often sparse, heterogeneous, and difficult to integrate. Although general-purpose Large Language Model (LLM) embeddings encode rich semantic information, they often lack the task-specific biomedical context needed for inference tasks such as computational drug repurposing. We introduce Contextualizing LLM Embeddings via Attention-based gRaph learning (CLEAR), a multimodal representation-fusion framework that aligns LLM embeddings with the topological structure of a context-specific Knowledge Graph (KG). Across five benchmark datasets, CLEAR achieved state-of-the-art results, improving predictive performance (e.g., F1 score) by up to 30% over prior methods. We further applied CLEAR to identify FDA-approved drugs with potential for repurposing for ADRD, including Parkinson disease-related dementia and Lewy Body dementia. CLEAR learned a biologically coherent embedding space, prioritized leading ADRD drug candidates, and accurately summarized known therapeutic relationships for FDA-approved Alzheimer disease drugs. Overall, CLEAR shows that grounding biomedical LLM embeddings with context-specific KG signals can improve drug repurposing in data-sparse, real-world settings. GitHub: https://github.com/bozdaglab/CLEAR

bioinformatics↗

bioRxiv2026

DyGraphTrans: A temporal graph representation learning framework for modeling disease progression from Electronic Health Records

MotivationElectronic Health Records (EHRs) contain vast amounts of longitudinal patient medical history data, making them highly informative for early disease prediction. Numerous computational methods have been developed to leverage EHR data; however, many process multiple patient records simultaneously, resulting in high memory consumption and computational cost. Moreover, these models also often lack interpretability, limiting insight into the factors driving their predictions. Efficiently handling large-scale EHR data while maintaining predictive accuracy and interpretability therefore remains a critical challenge. To address this gap, we propose DyGraphTrans, a dynamic graph representation learning framework that represents patient EHR data as a sequence of temporal graphs. In this representation, nodes correspond to patients, node features encode temporal clinical attributes, and edges capture patient similarity. DyGraphTrans models both local temporal dependencies and long-range global trends, while a sliding-window mechanism reduces memory consumption without sacrificing essential temporal context. Unlike existing dynamic graph models, DyGraphTrans jointly captures patient similarity and temporal evolution in a memory-efficient and interpretable manner. ResultsWe evaluated DyGraphTrans on Alzheimers Disease Neuroimaging Initiative (ADNI) and National Alzheimers Coordinating Center (NACC) for disease progression prediction, as well as on the Medical Information Mart for Intensive Care (MIMIC-IV) dataset for early mortality prediction. We further assessed the model on multiple benchmark dynamic graph datasets to evaluate its generalizability. DyGraphTrans achieved strong predictive performance across diverse datasets. We also demonstrated interpretability of DyGraphTrans aligned with known clinical risk factors. The source code and datasets are available at https://github.com/bozdaglab/DyGraphTrans. ContactSerdar.Bozdag@unt.edu SupplementaryUploaded as an attachment.

bioinformatics↗

bioRxiv2026

MultiGEOmics: Graph-Based Integration of Multi-Omics via Biological Information Flows

Multi-omics datasets capture complementary aspects of biological systems and are central to modern machine learning applications in biology and medicine. Existing graph-based integration methods typically construct separate graphs for each omics type and focus primarily on intra-omic relationships. As a result, they often overlook cross-omics regulatory signals--bidirectional interactions across omics layers--that are critical for modeling complex cellular processes. A second major challenge is missing or incomplete omics data; many current approaches degrade substantially in performance or exclude patients lacking one or more omics modalities. To address these limitations, we introduce MultiGEOmics, an intermediate-level graph integration framework that explicitly incorporates regulatory signals across omics types during graph representation learning and models biologically inspired omics-specific and cross-omics dependencies. MultiGEOmics learns robust cross-omics embeddings that remain reliable even when some modalities are partially missing. We evaluated MultiGEOmics across eleven datasets spanning cancer and Alzheimers disease, under zero, moderate, and high missing-rate scenarios. MultiGEOmics consistently maintains strong predictive performance across all missing-data conditions while offering interpretability by identifying the most influential omics types and features for each prediction task. The source code and the documentation of MultiGEOmics are available at https://github.com/bozdaglab/MultiGEOmics.

bioinformatics↗

bioRxiv2026

scAURA: Alignment- and Uniformity-based Graph DebiasedContrastive Representation Architecture for Self-SupervisedClustering of Single-Cell Transcriptomics

Single-cell RNA sequencing (scRNA-seq) allows transcriptomic profiling at single-cell resolution, providing valuable insights into cellular diversity across tissues, developmental stages, and diseases. However, accurately identifying cell types remains challenging due to the high dimensionality, sparsity, and noise inherent in scRNA-seq data. To address these challenges in cell type identification in scRNA-seq data, we introduce scAURA (single cell Alignment- and Uniformity-based Graph Debiased Contrastive Representation Architecture), a unified framework that integrates graph debiased contrastive learning with self-supervised clustering. We evaluated scAURA on 18 real single-cell datasets collected from six sequencing platforms spanning diverse tissue and cell types in human and mouse. scAURA outperformed all state-of-the-art (SOTA) methods in nine and eight datasets in Adjusted Rand Index (ARI) and Normalized Mutual Information (NMI), respectively. On average, scAURA obtained average ranks of 2.28 (ARI) and 2.39 (NMI) across all 13 SOTA methods, demonstrating its consistent superiority across datasets. scAURA also exhibited strong robustness to dropout noise by maintaining stable clustering performance even under increasing sparsity levels. Furthermore, in an external single-cell Alzheimers disease dataset, scAURA accurately clustered different cell types, identified novel cell type-specific marker genes, and inferred their potential transcriptional regulators. The source code and datasets are available at https://github.com/bozdaglab/scAURA. ContactSerdar.Bozdag@unt.edu

bioinformatics↗

bioRxiv2025

A Systematic Fairness Evaluation of Racial Bias in Alzheimers Disease Diagnosis Using Machine Learning Models

Alzheimers disease (AD) is a major global health concern, expected to affect 12.7 million Americans by 2050. Machine learning (ML) algorithms have been developed for AD diagnosis and progression prediction, but the lack of racial/ethnic diversity in clinical datasets raises concerns about their generalizability across demographic groups, particularly underrepresented populations. Studies show ML algorithms inherit biases from data, leading to biased AD predictions. This study investigates the fairness of ML models in AD diagnosis. We hypothesize that models trained on a single racial/ethnic group perform well within that group but poorly in others. We employ feature selection and model training techniques to improve fairness. Our findings support our hypothesis that ML models trained on one group underperform on others. We also demonstrated that applying fairness techniques to ML models reduces their bias. This study highlights the need for racial/ethnic diversity in datasets and fair models for AD prediction.

bioinformatics↗

bioRxiv2025

A network model for patient-derived drug response in breast cancer integrating multi-omics datasets.

The widespread availability of multi-omics tumor profiling has enabled detailed molecular characterization of individual tumors, paving the way for more effective, less toxic, and patient-specific therapies. However, widespread compound screens in human patients are constrained by ethical and logistical challenges, underscoring the need for computational models capable of predicting in vivo drug response. Here, we introduce PDDRNet-MH, a multiplex heterogeneous network-based framework that integrates genomic, transcriptomic, and epigenomic tumor profiles with drug chemical structures, pharmacological activity, and side-effect data to infer personalized drug responses. PDDRNet-MH constructs an integrated network connecting patients, cell lines, drugs, and genes, with each component represented by four biologically and pharmacologically informed similarity layers. This design enables the systematic propagation of drug-biomarker associations across modalities. We applied PDDRNet-MH to breast cancer patient data from The Cancer Genome Atlas and benchmarked its predictive performance against state-of-the-art methods on eleven FDA-approved breast cancer drugs. PDDRNet-MH achieved consistently high accuracy, with perfect prediction scores for gemcitabine and vinorelbine (Area Under the Receiver Operating Characteristic Curve [AUC-ROC] = 1.00; Area Under the Precision-Recall Curve [AUC-PR] = 1.00) and near-perfect scores for methotrexate and zoledronate (AUC-ROC = 0.95; AUC-PR = 0.99), demonstrating its ability to robustly distinguish sensitive from resistant patients. Biologically, PDDRNet-MH accurately prioritized established clinical biomarkers, including HER2 (ERBB2) for lapatinib and BRCA1/2 for doxorubicin and cyclophosphamide. Beyond known associations, the model identified additional genes within the HER2 amplicon on chromosome 17q12, including STARD3, MIEN1, and PPP1R1B, whose amplification was significantly associated with elevated drug response scores, suggesting potential roles in HER2-targeted therapy. These findings highlight the ability of PDDRNet-MH to recover and extend clinically relevant drug-biomarker associations, supporting its utility in guiding precision oncology.

cancer biology↗

bioRxiv2025

ProtFun: A Protein Function Prediction Model Using Graph Attention Networks with a Protein Large Language Model

Understanding protein functions facilitates the identification of the underlying causes of many diseases and guides the research for discovering new therapeutic targets and medications. With the advancement of high throughput technologies, obtaining novel protein sequences has been a routine process. However, determining protein functions experimentally is cost- and labor-prohibitive. Therefore, it is crucial to develop computational methods for automatic protein function prediction. In this study, we propose a multi-modal deep learning architecture called ProtFun to predict protein functions. ProtFun integrates protein large language model (LLM) embeddings as node features in a protein family network. Employing graph attention networks (GAT) on this protein family network, ProtFun learns protein embeddings, which are integrated with protein signature representations from InterPro to train a protein function prediction model. We evaluated our architecture using three benchmark datasets. Our results showed that our proposed approach outperformed current state-of-the-art methods for most cases. An ablation study also highlighted the importance of different components of ProtFun. The data and source code of ProtFun is available at https://github.com/bozdaglab/ProtFun under Creative Commons Attribution Non Commercial 4.0 International Public License.

bioinformatics↗

bioRxiv2025

Integrated single cell spatial multi-omics landscape of WHO grades 2-4 diffuse gliomas identifies locoregional metabolomic regulators of glioma growth

Diffuse infiltrating gliomas are aggressive tumors of the central nervous system driven by intra-tumoral heterogeneity and aberrant normal-tumor cell-cell interactions. Grade specific and locoregional metabolic dependencies driving aberrant cell-states linked to treatment resistance, seizures and infiltration of gliomas remain elusive. Here, we applied spatial transcriptomics (stRNAseq), imaging mass cytometry (IMC) and mass spectrometry imaging (MSI; metabolites, peptides and glycans) to the core and edge tumor tissue from patients with World Health Organization (WHO) grades 2-4 diffuse infiltrating gliomas including isocitrate dehydrogenase (IDH) mutant oligodendrogliomas (WHO Grades 2 and 3) and IDH wildtype astrocytomas including anaplastic astrocytoma (prior 2016 WHO histological grade 3) and glioblastoma (GBM, WHO grade 4) stRNAseq identified regions-specific differentially expressed genes with significant overall survival implications particularly in IDH wildtype GBM. Integration of stRNA seq and MSI-derived metabolite expression demonstrated enrichment of L-glutamine in SOX4+ Neural progenitor-like (NPC-like) cells and DL-dopamine in GPNMB+ Mesenchymal-like (MES-like) GBM cells at the tumor edge relative to the core. Our results uncover clinically relevant and locoregional cell state-specific metabolites that may contribute to GBM proliferation, infiltration and seizures. This comprehensive pan-diffuse infiltrating glioma multi-omics study could serve as a resource for uncovering region-specific metabolic vulnerabilities encompassing metabolites, glycans and peptides within transcriptionally defined cell states across WHO 2-4 diffuse glioma.

cancer biology↗

bioRxiv2025

Top-DTI: Integrating Topological Deep Learning and Large Language Models for Drug Target Interaction Prediction

MotivationThe accurate prediction of drug-target interactions (DTI) is a crucial step in drug discovery, providing a foundation for identifying novel therapeutics. Traditional drug development is both costly and time-consuming, often spanning over a decade. Computational approaches help narrow the pool of compound candidates, offering significant starting points for experimental validation. In this study, we propose Top-DTI framework for predicting DTI by integrating topological data analysis (TDA) with large language models (LLMs). Top-DTI leverages persistent homology to extract topological features from protein contact maps and drug molecular images. Simultaneously, protein and drug LLMs generate semantically rich embeddings that capture sequential and contextual information from protein sequences and drug SMILES strings. By combining these complementary features, Top-DTI enhances predictive performance and robustness. ResultsExperimental results on the public BioSNAP and Human DTI benchmark datasets demonstrate that the proposed Top-DTI model outperforms state-of-the-art approaches across multiple evaluation metrics, including AUROC, AUPRC, sensitivity, and specificity. Furthermore, the Top-DTI model achieves superior performance in the challenging cold-split scenario, where the test and validation sets contain drugs or targets absent from the training set. This setting simulates real-world scenarios and highlights the robustness of the model. Notably, incorporating topological features alongside LLM embeddings significantly improves predictive performance, underscoring the value of integrating structural and sequence-based representations. AvailabilityThe data and source code of Top-DTI is available at https://github.com/bozdaglab/Top_DTI under Creative Commons Attribution Non Commercial 4.0 International Public License.

bioinformatics↗

bioRxiv2024

Identifying Alzheimer's disease-associated genes using PhenoGeneRanker

Alzheimers disease (AD) is a neurogenerative disease that affects millions worldwide with no effective treatment. Several studies have been conducted to decipher to genomic underpinnings of AD. Due to its complex nature, many genes have been found to be associated with AD. Despite these findings, the pathophysiology of the disease is still elusive. To discover new putative AD-associated genes, in this study, we integrated multimodal gene and phenotype datasets of AD using network biology methods to prioritize potential AD-related genes. We constructed a multiplex heterogeneous network composed of patient and gene similarity networks utilizing phenotypic and omics datasets of AD patients from the Alzheimers Disease Neuroimaging Initiative (ADNI) database. We applied PhenoGeneRanker to traverse this network to discover potential AD-associated genes. To assess the impact of each network layer and seed gene, we also run PhenoGeneRanker on different variants of the network and seed genes. Our results showed that top-ranked genes captured several known AD-related genes and were enriched in Gene Ontology (GO) terms related to AD. We also observed that several top-ranked genes that are not in AD-associated gene list had literature supporting their potential relevance to AD.

bioinformatics↗

bioRxiv2024

Early changes in microRNA expression in Arabidopsis plants infected with the fungal pathogen Fusarium graminearum

Plants respond to biotic stressors by modulating various processes in an attempt to limit the attack by a pathogen or herbivore. Triggering these different defense processes requires orchestration of a network of proteins and RNA molecules that includes microRNAs (miRNAs). These short RNA molecules (20-22 nucleotides) have been shown to be important players in the early responses of plants to stresses because they can rapidly regulate the expression levels of a network of downstream genes. The ascomycete Fusarium graminearum is an important fungal pathogen that causes significant losses in cereal crops worldwide. Using the well-characterized Fusarium-Arabidopsis pathosystem, we investigated how plants change expression of their miRNAs globally during the early stages of infection by F. graminearum. In addition to miRNAs that have been previously implicated in stress responses, we have also identified evolutionarily young miRNAs whose levels change significantly in response to fungal infection. Some of these young miRNAs have homologs present in cereals. Thus, manipulating expression of these miRNAs may provide a unique path toward development of plants with increased resistance to fungal pathogens.

plant biology↗

bioRxiv2023

PVTAD: ALZHEIMERS DISEASE DIAGNOSIS USING PYRAMID VISION TRANSFORMER APPLIED TO WHITE MATTER OF T1-WEIGHTED STRUCTURAL MRI DATA

Alzheimers disease (AD) is a neurodegenerative disorder, and timely diagnosis is crucial for early interventions. AD is known to have disruptive local and global brain neural connections that may be instrumental in understanding and extracting specific biomarkers. Previous machine-learning approaches are mostly based on convolutional neural network (CNN) and standard vision transformer (ViT) models which may not sufficiently capture the multidimensional local and global patterns that may be indicative of AD. Therefore, in this paper, we propose a novel approach called PVTAD to classify AD and cognitively normal (CN) cases using pretrained pyramid vision transformer (PVT) and white matter (WM) of T1-weighted structural MRI (sMRI) data. Our approach combines the advantages of CNN and standard ViT to extract both local and global features indicative of AD from the WM coronal middle slices. We performed experiments on subjects with T1-weighed MPRAGE sMRI scans from the ADNI dataset. Our results demonstrate that the PVTAD achieves an average accuracy of 97.7% and F1-score of 97.6%, outperforming the single and parallel CNN and standard ViT architectures based on sMRI data for AD vs. CN classification.

bioinformatics↗

bioRxiv2023

PPAD: A deep learning architecture to predict progression of Alzheimer's disease

Alzheimers disease (AD) is a neurodegenerative disease that affects millions of people worldwide. Mild cognitive impairment (MCI) is an intermediary stage between cognitively normal (CN) state and AD. Not all people who have MCI convert to AD. The diagnosis of AD is made after significant symptoms of dementia such as short-term memory loss are already present. Since AD is currently an irreversible disease, diagnosis at the onset of disease brings a huge burden on patients, their caregivers, and the healthcare sector. Thus, there is a crucial need to develop methods for the early prediction AD for patients who have MCI. Recurrent Neural Networks (RNN) have been successfully used to handle Electronic Health Records (EHR) for predicting conversion from MCI to AD. However, RNN ignores irregular time intervals between successive events which occurs common in EHR data. In this study, we propose two deep learning architectures based on RNN, namely Predicting Progression of Alzheimers Disease (PPAD) and PPAD-Autoencoder (PPAD-AE). PPAD and PPAD-AE are designed for early predicting conversion from MCI to AD at the next visit and multiple visits ahead for patients, respectively. To minimize the effect of the irregular time intervals between visits, we propose using age in each visit as an indicator of time change between successive visits. Our experimental results conducted on Alzheimers Disease Neuroimaging Initiative (ADNI) and National Alzheimers Coordinating Center (NACC) datasets showed that our proposed models outperformed all baseline models for most prediction scenarios in terms of F2 and sensitivity. We also observed that the age feature was one of top features and was able to address irregular time interval problem.

bioinformatics↗

bioRxiv2022

Finding the best cell lines across pan-cancer to use in pre-clinical research as a proxy for patient tumor samples considering immune cells, multi-omics, and cancer pathways

In pre-clinical trials of anti-cancer drugs, cell lines are utilized as a model for patient tumor samples to understand the response of drugs. However, in vitro culture of cell lines, in general, alters the biology of the cell lines and likely gives rise to systematic differences from the tumor samples genomic profiles; hence the drug response of cell lines may deviate from actual patients drug response. In this study, we computed a similarity score for the selection of cell lines depicting the close and far resemblance to patient tumor samples in twenty-two different cancer types at genetic, genomic, and epigenetic levels integrating multi-omics datasets. We also considered the presence of immune cells in tumor samples and cancer-related biological pathways in this score which aids personalized medicine research in cancer. We showed that based on these similarity scores cell lines were able to recapitulate the drug response of patient tumor samples for several FDA-approved cancer drugs in multiple cancer types. Based on these scores, several of the high-rank cell lines were shown to have a close likeness to the corresponding tumor type in previously reported in vitro experiments.

bioinformatics↗

bioRxiv2022

NRPreTo: A Machine Learning Based Nuclear Receptor and Subfamily Prediction Tool

The Nuclear Receptor (NR) superfamily includes phylogenetically related ligand-activated proteins, which play a key role in various cellular activities. NR proteins are subdivided into seven subfamilies based on their function, mechanism, and nature of the interacting ligand. Developing robust tools to identify NR could give insights into their functional relationships and involvement in disease pathways. Existing NR prediction tools only use a few types of sequence-based features and are tested on relatively similar independent datasets; thus, they may suffer from overfitting when extended to new genera of sequences. To address this problem, we developed Nuclear Receptor Prediction Tool (NRPreTo); a two-level NR prediction tool with a unique training approach where in addition to the sequence-based features used by existing NR prediction tools, six additional feature groups depicting various physiochemical, structural and evolutionary features of proteins were utilized. The first level of NRPreTo allows for the successful prediction of a query protein as NR or non-NR, and further subclassifies the protein into one of the seven NR subfamilies in the second level. We developed Random Forest classifiers to test on benchmark datasets, as well as the entire human protein datasets from RefSeq and Human Protein Reference Database (HPRD). We observed that using additional feature groups improved performance. We also observed that NRPreTo achieved high performance on the external datasets and predicted 59 novel NRs in the human proteome. The source code of NRPreTo is publicly available at https://github.com/bozdaglab/NRPreTo.

systems biology↗

bioRxiv2022

SUPREME: A cancer subtype prediction methodology integrating multiomics data using Graph Convolutional Neural Network

To pave the road towards precision medicine in cancer, patients with highly similar biology ought to be grouped into the same cancer subtypes. Utilizing high-dimensional multiomics datasets, several integrative computational approaches have been developed to uncover cancer subtypes. Recently, Graph Neural Networks (GNNs) was discovered to learn node embeddings while utilizing node features and node associations at the same time on graph-structured data. Although there are some commonly used architectures such as Graph Convolutional Network (GCN) for cancer subtype prediction, the existing prediction tools have some limitations in leveraging those architectures with multiomics integration on multiple networks. Addressing them, we developed SUPREME (a subtype prediction methodology) by comprehensively analyzing multiomics data and associations between patients with graph convolutions on multiple patient similarity networks. Unlike the existing tools, SUPREME generates patient embeddings from patient similarity networks, on which it utilizes all the multiomics features. In addition, SUPREME integrates all the possible combinations of embeddings with the raw multiomics features to capture the complementary signals. Extensive evaluation of all combinations makes SUPREME interpretable in terms of utilized networks and features. On three different datasets from The Cancer Genome Atlas (TCGA), Molecular Taxonomy of Breast Cancer International Consortium (METABRIC), and both combined, our method significantly outperformed other integrative cancer (sub)type prediction tools and baseline methods, with overall consistent results. SUPREME-inferred subtypes had significant survival differences, mostly having more significance than ground truth (PAM50) labels, and outperformed nine cancer subtype differentiating tools and baseline methods. These results suggest that with proper utilization of multiple datatypes and patient associations, SUPREME could demystify the undiscovered characteristics in cancer subtypes that cause significant survival differences and could improve the ground truth label, which depends mainly on a single datatype. Source code for our tool is publicly available at https://github.com/bozdaglab/SUPREME.

bioinformatics↗

bioRxiv2020

Inferring competing endogenous RNA (ceRNA) interactions in cancer

To understand driving biological factors for complex diseases like cancer, regulatory circuity of genes needs to be discovered. Recently, a new gene regulation mechanism called competing endogenous RNA (ceRNA) interactions has been discovered. Certain genes targeted by common microRNAs (miRNAs) "compete" for these miRNAs, thereby regulate each other by making others free from miRNA regulation. Several computational tools have been published to infer ceRNA networks. In most existing tools, however, expression abundance sufficiency, collective regulation, and groupwise effect of ceRNAs are not considered. In this study, we developed a computational tool named Crinet to infer genome-wide ceRNA networks addressing critical drawbacks. Crinet considers all mRNAs, lncRNAs, and pseudogenes as potential ceRNAs and incorporates a network deconvolution method to exclude the spurious ceRNA pairs. We tested Crinet on breast cancer data in TCGA. Crinet inferred reproducible ceRNA interactions and groups, which were significantly enriched in the cancer-related genes and processes. We validated the selected miRNA-target interactions with the protein expression-based benchmarks and also evaluated the inferred ceRNA interactions predicting gene expression change in knockdown assays. The hub genes in the inferred ceRNA network included known suppressor/oncogene lncRNAs in breast cancer showing the importance of non-coding RNAs inclusion for ceRNA inference. The source code of Crinet could be accessed on Github at https://github.com/bozdaglab/crinet.

bioinformatics↗

bioRxiv2020

Gene Embeddings of Complex network (GECo) and hypertension disease gene classification

Complex diseases such as hypertension, cancer, and diabetes cause nearly 70% of the deaths in the U.S. and involve multiple genes and their interactions with environmental factors. Therefore, identification of genetic factors to understand and decrease the morbidity and mortality from complex diseases is an important and challenging task. With the generation of an unprecedented amount of multi-omics datasets, network-based methods have become popular to represent the multilayered complex molecular interactions. Particularly node embeddings, the low-dimensional representations of nodes in a network are utilized for gene function prediction. Integrated network analysis of multi-omics data alleviates the issues related to missing data and lack of context-specific datasets. Most of the node embedding methods, however, are unable to integrate multiple types of datasets from genes and phenotypes. To address this limitation, we developed a node embedding algorithm called Node Embeddings of Complex networks (NECo) that can utilize multilayered heterogeneous networks of genes and phenotypes. We evaluated the performance of NECo using genotypic and phenotypic datasets from rat (Rattus norvegicus) disease models to classify hypertension disease-related genes. Our method significantly outperformed the state-of-the-art node embedding methods, with AUC of 94.97% compared 85.98% in the second-best performer, and predicted genes not previously implicated in hypertension. Availability and implementationThe source code is available on GitHub at https://github.com/bozdaglab/NECo.

bioinformatics↗