MOTIVATION: Developing computational tools for integrative analysis across multiple types of omics data has been of immense importance in cancer molecular biology and precision medicine research. While recent advancements have yielded integrative prediction solutions for multi-omics data, these methods lack a comprehensive and cohesive understanding of the rationale behind their specific predictions. To shed light on personalized medicine and unravel previously unknown characteristics within integrative analysis of multi-omics data, we introduce a novel integrative neural network approach for cancer molecular subtype and biomedical classification applications, named Integrative Graph Convolutional Networks (IGCN). RESULTS: To demonstrate the superiority of IGCN, we compare its performance with other state-of-the-art approaches across different cancer subtype and biomedical classification tasks. Our experimental results show that our proposed model outperforms the state-of-the-art and baseline methods. IGCN identifies which types of omics data receive more emphasis for each patient when predicting a specific class. Additionally, IGCN has the capability to pinpoint significant biomarkers from a range of omics data types. AVAILABILITY AND IMPLEMENTATION: The source code is available at https://github.com/bozdaglab/IGCN.

MOTIVATION: The 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 a 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. RESULTS: Experimental 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. AVAILABILITY AND IMPLEMENTATION: The data and source code of Top-DTI are available at https://github.com/bozdaglab/Top_DTI under the Creative Commons Attribution NonCommercial 4.0 International Public License.

MOTIVATION: Electronic health records (EHRs) represent a comprehensive resource of a patient's medical history. EHRs are essential for utilizing advanced technologies such as deep learning (DL), enabling healthcare providers to analyze extensive data, extract valuable insights, and make precise and data-driven clinical decisions. DL methods such as recurrent neural networks (RNN) have been utilized to analyze EHR to model disease progression and predict diagnosis. However, these methods do not address some inherent irregularities in EHR data such as irregular time intervals between clinical visits. Furthermore, most DL models are not interpretable. In this study, we propose two interpretable DL architectures based on RNN, namely time-aware RNN (TA-RNN) and TA-RNN-autoencoder (TA-RNN-AE) to predict patient's clinical outcome in EHR at the next visit and multiple visits ahead, respectively. To mitigate the impact of irregular time intervals, we propose incorporating time embedding of the elapsed times between visits. For interpretability, we propose employing a dual-level attention mechanism that operates between visits and features within each visit. RESULTS: The results of the experiments conducted on Alzheimer's Disease Neuroimaging Initiative (ADNI) and National Alzheimer's Coordinating Center (NACC) datasets indicated the superior performance of proposed models for predicting Alzheimer's Disease (AD) compared to state-of-the-art and baseline approaches based on F2 and sensitivity. Additionally, TA-RNN showed superior performance on the Medical Information Mart for Intensive Care (MIMIC-III) dataset for mortality prediction. In our ablation study, we observed enhanced predictive performance by incorporating time embedding and attention mechanisms. Finally, investigating attention weights helped identify influential visits and features in predictions. AVAILABILITY AND IMPLEMENTATION: https://github.com/bozdaglab/TA-RNN.

MOTIVATION: Alzheimer's disease (AD) is a neurodegenerative disease that affects millions of people worldwide. Mild cognitive impairment (MCI) is an intermediary stage between cognitively normal 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 the 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 electronic health record data. In this study, we propose two deep learning architectures based on RNN, namely Predicting Progression of Alzheimer's Disease (PPAD) and PPAD-Autoencoder. PPAD and PPAD-Autoencoder 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. RESULTS: Our experimental results conducted on Alzheimer's Disease Neuroimaging Initiative and National Alzheimer's Coordinating Center 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. AVAILABILITY AND IMPLEMENTATION: https://github.com/bozdaglab/PPAD.

Autism spectrum disorder (ASD) affects large number of children and adults in the US, and worldwide. Early and quick diagnosis of ASD can improve the quality of life significantly both for patients and their families. Prior research provides strong evidence that structural and functional magnetic resonance imaging (MRI) data collected from individuals with ASD exhibit distinguishing characteristics that differ in local and global, spatial and temporal neural patterns of the brain - and therefore can be used for diagnostic purposes for various mental disorders. However, the data from MRI are high-dimensional and advanced methods are needed to make sense out of these datasets. In this paper, we present a novel model based on graph convolutional network (GCN) that can utilize resting state fMRI (rs-fMRI) data to classify ASD subjects from health controls (HC). In addition to using the graph from traditional correlation matrices, our proposed GCN model incorporates graphlet topological counting as one of the training features. Our results show that graphlets can preserve the topological information of the graphs obtained from fMRI data. Combined with our GCN, the graphlets retain enough topological information to differentiate between the ASD and HC. Our proposed model gives an average accuracy of 64.27% on the whole ABIDE-I data sets (1035 subjects) and highest site-specific accuracy of 75.9%, which is comparable to other state-of-the-art methods - while potentially open to being more interpretable.

Uncovering genotype-phenotype relationships is a fundamental challenge in genomics. Gene prioritization is an important step for this endeavor to make a short manageable list from a list of thousands of genes coming from high-throughput studies. Network propagation methods are promising and state of the art methods for gene prioritization based on the premise that functionally related genes tend to be close to each other in the biological networks. Recently, we introduced PhenoGeneRanker, a network-propagation algorithm for multiplex heterogeneous networks. PhenoGeneRanker allows multi-layer gene and phenotype networks. It also calculates empirical p values of gene and phenotype ranks using random stratified sampling of seeds of genes and phenotypes based on their connectivity degree in the network. In this study, we introduce the PhenoGeneRanker Bioconductor package and its application to multi-omics rat genome datasets to rank hypertension disease-related genes and strains. We showed that PhenoGeneRanker performed better to rank hypertension disease-related genes using multiplex gene networks than aggregated gene networks. We also showed that PhenoGeneRanker performed better to rank hypertension disease-related strains using multiplex phenotype network than single or aggregated phenotype networks. We performed a rigorous hyperparameter analysis and, finally showed that Gene Ontology (GO) enrichment of statistically significant top-ranked genes resulted in hypertension disease-related GO terms.

Abstract 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 implementation The source code is available on GitHub at https://github.com/bozdaglab/NECo .

BACKGROUND: RNA-seq, wherein RNA transcripts expressed in a sample are sequenced and quantified, has become a widely used technique to study disease and development. With RNA-seq, transcription abundance can be measured, differential expression genes between groups and functional enrichment of those genes can be computed. However, biological insights from RNA-seq are often limited by computational analysis and the enormous volume of resulting data, preventing facile and meaningful review and interpretation of gene expression profiles. Particularly, in cases where the samples under study exhibit uncontrolled variation, deeper analysis of functional enrichment would be necessary to visualize samples' gene expression activity under each biological function. RESULTS: We developed a Bioconductor package rgsepd that streamlines RNA-seq data analysis by wrapping commonly used tools DESeq2 and GOSeq in a user-friendly interface and performs a gene-subset linear projection to cluster heterogeneous samples by Gene Ontology (GO) terms. Rgsepd computes significantly enriched GO terms for each experimental condition and generates multidimensional projection plots highlighting how each predefined gene set's multidimensional expression may delineate samples. CONCLUSIONS: The rgsepd serves to automate differential expression, functional annotation, and exploratory data analyses to highlight subtle expression differences among samples based on each significant biological function.

MicroRNAs (miRNAs) inhibit expression of target genes by binding to their RNA transcripts. It has been recently shown that RNA transcripts targeted by the same miRNA could "compete" for the miRNA molecules and thereby indirectly regulate each other. Experimental evidence has suggested that the aberration of such miRNA-mediated interaction between RNAs-called competing endogenous RNA (ceRNA) interaction-can play important roles in tumorigenesis. Given the difficulty of deciphering context-specific miRNA binding, and the existence of various gene regulatory factors such as DNA methylation and copy number alteration, inferring context-specific ceRNA interactions accurately is a computationally challenging task. Here we propose a computational method called Cancerin to identify cancer-associated ceRNA interactions. Cancerin incorporates DNA methylation, copy number alteration, gene and miRNA expression datasets to construct cancer-specific ceRNA networks. We applied Cancerin to three cancer datasets from the Cancer Genome Atlas (TCGA) project. Our results indicated that ceRNAs were enriched with cancer-related genes, and ceRNA modules in the inferred ceRNA networks were involved in cancer-associated biological processes. Using LINCS-L1000 shRNA-mediated gene knockdown experiment in breast cancer cell line to assess accuracy, Cancerin was able to predict expression outcome of ceRNA genes with high accuracy.

For the vast majority of species - including many economically or ecologically important organisms, progress in biological research is hampered due to the lack of a reference genome sequence. Despite recent advances in sequencing technologies, several factors still limit the availability of such a critical resource. At the same time, many research groups and international consortia have already produced BAC libraries and physical maps and now are in a position to proceed with the development of whole-genome sequences organized around a physical map anchored to a genetic map. We propose a BAC-by-BAC sequencing protocol that combines combinatorial pooling design and second-generation sequencing technology to efficiently approach denovo selective genome sequencing. We show that combinatorial pooling is a cost-effective and practical alternative to exhaustive DNA barcoding when preparing sequencing libraries for hundreds or thousands of DNA samples, such as in this case gene-bearing minimum-tiling-path BAC clones. The novelty of the protocol hinges on the computational ability to efficiently compare hundred millions of short reads and assign them to the correct BAC clones (deconvolution) so that the assembly can be carried out clone-by-clone. Experimental results on simulated data for the rice genome show that the deconvolution is very accurate, and the resulting BAC assemblies have high quality. Results on real data for a gene-rich subset of the barley genome confirm that the deconvolution is accurate and the BAC assemblies have good quality. While our method cannot provide the level of completeness that one would achieve with a comprehensive whole-genome sequencing project, we show that it is quite successful in reconstructing the gene sequences within BACs. In the case of plants such as barley, this level of sequence knowledge is sufficient to support critical end-point objectives such as map-based cloning and marker-assisted breeding.

The problem of computing the minimum tiling path (MTP) from a set of clones arranged in a physical map is a cornerstone of hierarchical (clone-by-clone) genome sequencing projects. We formulate this problem in a graph theoretical framework, and then solve by a combination of minimum hitting set and minimum spanning tree algorithms. The tool implementing this strategy, called FMTP, shows improved performance compared to the widely used software FPC. When we execute FMTP and FPC on the same physical map, the MTP produced by FMTP covers a higher portion of the genome, and uses a smaller number of clones. For instance, on the rice genome the MTP produced by our tool would reduce by about 11 percent the cost of a clone-by-clone sequencing project. Source code, benchmark data sets, and documentation of FMTP are freely available at >http://code.google.com/p/fingerprint-based-minimal-tiling-path/ under MIT license.

MOTIVATION: In order to construct gene regulatory networks of higher organisms from gene expression and promoter sequence data efficiently, we developed FastMEDUSA. In this parallelized version of the regulatory network-modeling tool MEDUSA, expression and sequence data are shared among a user-defined number of processors on a single multi-core machine or cluster. Our results show that FastMEDUSA allows a more efficient utilization of computational resources. While the determination of a regulatory network of brain tumor in Homo sapiens takes 12 days with MEDUSA, FastMEDUSA obtained the same results in 6 h by utilizing 100 processors. AVAILABILITY: Source code and documentation of FastMEDUSA are available at https://wiki.nci.nih.gov/display/NOBbioinf/FastMEDUSA

BACKGROUND: Physical maps have been historically one of the cornerstones of genome sequencing and map-based cloning strategies. They also support marker assisted breeding and EST mapping. The problem of building a high quality physical map is computationally challenging due to unavoidable noise in the input fingerprint data. RESULTS: We propose a novel compartmentalized method for the assembly of high quality physical maps from fingerprinted clones. The knowledge of genetic markers enables us to group clones into clusters so that clones in the same cluster are more likely to overlap. For each cluster of clones, a local physical map is first constructed using FingerPrinted Contigs (FPC). Then, all the individual maps are carefully merged into the final physical map. Experimental results on the genomes of rice and barley demonstrate that the compartmentalized assembly produces significantly more accurate maps, and that it can detect and isolate clones that would induce "chimeric" contigs if used in the final assembly. CONCLUSION: The software is available for download at http://www.cs.ucr.edu/~sbozdag/assembler/

We propose a novel compartmentalized method for the assembly of physical maps from fingerprinted clones. Our assembler exploits the presence of genetic markers at the global level to improve the accuracy of the assembly. Experimental results on the genome of rice and barley demonstrate that the compartmentalized assembler produces significantly more accurate maps, and that it can detect and isolate clones that induce chimeric contigs.