The value of prospective metabolomic susceptibility endotypes: broad applicability for infectious diseases

BACKGROUND

As new infectious diseases (ID) emerge and others continue to mutate, there remains an imminent threat, especially for vulnerable individuals. Yet no generalizable framework exists to identify the at-risk group prior to infection. Metabolomics has the advantage of capturing the existing physiologic state, unobserved via current clinical measures. Furthermore, metabolomics profiling during acute disease can be influenced by confounding factors such as indications, medical treatments, and lifestyles.

METHODS

We employed metabolomic profiling to cluster infection-free individuals and assessed their relationship with COVID severity and influenza incidence/recurrence.

FINDINGS

We identified a metabolomic susceptibility endotype that was strongly associated with both severe COVID (ORICUadmission = 6.7, p-value = 1.2 × 10-08, ORmortality = 4.7, p-value = 1.6 × 10-04) and influenza (ORincidence = 2.9; p-values = 2.2 × 10-4, βrecurrence = 1.03; p-value = 5.1 × 10-3). We observed similar severity associations when recapitulating this susceptibility endotype using metabolomics from individuals during and after acute COVID infection. We demonstrate the value of using metabolomic endotyping to identify a metabolically susceptible group for two-and potentially more-IDs that are driven by increases in specific amino acids, including microbial-related metabolites such as tryptophan, bile acids, histidine, polyamine, phenylalanine, and tyrosine metabolism, as well as carbohydrates involved in glycolysis.

INTERPRETATIONS

These metabolites may be identified prior to infection to enable protective measures for these individuals.

FUNDING

The Longitudinal EMR and Omics COVID-19 Cohort (LEOCC) and metabolomic profiling were supported by the National Heart, Lung, and Blood Institute and the Intramural Research Program of the National Center for Advancing Translational Sciences, National Institutes of Health.

Integrated unbiased multiomics defines disease-independent placental clusters in common obstetrical syndromes

BACKGROUND

Placental dysfunction, a root cause of common syndromes affecting human pregnancy, such as preeclampsia (PE), fetal growth restriction (FGR), and spontaneous preterm delivery (sPTD), remains poorly defined. These common, yet clinically disparate obstetrical syndromes share similar placental histopathologic patterns, while individuals within each syndrome present distinct molecular changes, challenging our understanding and hindering our ability to prevent and treat these syndromes.

METHODS

Using our extensive biobank, we identified women with severe PE (n = 75), FGR (n = 40), FGR with a hypertensive disorder (FGR + HDP; n = 33), sPTD (n = 72), and two uncomplicated control groups, term (n = 113), and preterm without PE, FGR, or sPTD (n = 16). We used placental biopsies for transcriptomics, proteomics, metabolomics data, and histological evaluation. After conventional pairwise comparison, we deployed an unbiased, AI-based similarity network fusion (SNF) to integrate the datatypes and identify omics-defined placental clusters. We used Bayesian model selection to compare the association between the histopathological features and disease conditions vs SNF clusters.

RESULTS

Pairwise, disease-based comparisons exhibited relatively few differences, likely reflecting the heterogeneity of the clinical syndromes. Therefore, we deployed the unbiased, omics-based SNF method. Our analysis resulted in four distinct clusters, which were mostly dominated by a specific syndrome. Notably, the cluster dominated by early-onset PE exhibited strong placental dysfunction patterns, with weaker injury patterns in the cluster dominated by sPTD. The SNF-defined clusters exhibited better correlation with the histopathology than the predefined disease groups.

CONCLUSIONS

Our results demonstrate that integrated omics-based SNF distinctively reclassifies placental dysfunction patterns underlying the common obstetrical syndromes, improves our understanding of the pathological processes, and could promote a search for more personalized interventions.

Identifying chronic obstructive pulmonary disease from integrative omics and clustering in lung tissue

BACKGROUND

Chronic obstructive pulmonary disease (COPD) is a highly morbid and heterogenous disease. While COPD is defined by spirometry, many COPD characteristics are seen in cigarette smokers with normal spirometry. The extent to which COPD and COPD heterogeneity is captured in omics of lung tissue is not known.

METHODS

We clustered gene expression and methylation data in 78 lung tissue samples from former smokers with normal lung function or severe COPD. We applied two integrative omics clustering methods: (1) Similarity Network Fusion (SNF) and (2) Entropy-Based Consensus Clustering (ECC).

RESULTS

SNF clusters were not significantly different by the percentage of COPD cases (48.8% vs. 68.6%, p = 0.13), though were different according to median forced expiratory volume in one second (FEV1) % predicted (82 vs. 31, p = 0.017). In contrast, the ECC clusters showed stronger evidence of separation by COPD case status (48.2% vs. 81.8%, p = 0.013) and similar stratification by median FEV1% predicted (82 vs. 30.5, p = 0.0059). ECC clusters using both gene expression and methylation were identical to the ECC clustering solution generated using methylation data alone. Both methods selected clusters with differentially expressed transcripts enriched for interleukin signaling and immunoregulatory interactions between lymphoid and non-lymphoid cells.

CONCLUSIONS

Unsupervised clustering analysis from integrated gene expression and methylation data in lung tissue resulted in clusters with modest concordance with COPD, though were enriched in pathways potentially contributing to COPD-related pathology and heterogeneity.

Single-nucleus gene and gene set expression-based similarity network fusion identifies autism molecular subtypes

BACKGROUND

Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder that is highly phenotypically and genetically heterogeneous. With the accumulation of biological sequencing data, more and more studies shift to molecular subtype-first approach, from identifying molecular subtypes based on genetic and molecular data to linking molecular subtypes with clinical manifestation, which can reduce heterogeneity before phenotypic profiling.

RESULTS

In this study, we perform similarity network fusion to integrate gene and gene set expression data of multiple human brain cell types for ASD molecular subtype identification. Then we apply subtype-specific differential gene and gene set expression analyses to study expression patterns specific to molecular subtypes in each cell type. To demonstrate the biological and practical significance, we analyze the molecular subtypes, investigate their correlation with ASD clinical phenotype, and construct ASD molecular subtype prediction models.

CONCLUSIONS

The identified molecular subtype-specific gene and gene set expression may be used to differentiate ASD molecular subtypes, facilitating the diagnosis and treatment of ASD. Our method provides an analytical pipeline for the identification of molecular subtypes and even disease subtypes of complex disorders.

BRWMC: Predicting lncRNA-disease associations based on bi-random walk and matrix completion on disease and lncRNA networks

Many experiments have proved that long non-coding RNAs (lncRNAs) in humans have been implicated in disease development. The prediction of lncRNA-disease association is essential in promoting disease treatment and drug development. It is time-consuming and laborious to explore the relationship between lncRNA and diseases in the laboratory. The computation-based approach has clear advantages and has become a promising research direction. This paper proposes a new lncRNA disease association prediction algorithm BRWMC. Firstly, BRWMC constructed several lncRNA (disease) similarity networks based on different measurement angles and fused them into an integrated similarity network by similarity network fusion (SNF). In addition, the random walk method is used to preprocess the known lncRNA-disease association matrix and calculate the estimated scores of potential lncRNA-disease associations. Finally, the matrix completion method accurately predicts the potential lncRNA-disease associations. Under the framework of leave-one-out cross-validation and 5-fold cross-validation, the AUC values obtained by BRWMC are 0.9610 and 0.9739, respectively. In addition, case studies of three common diseases show that BRWMC is a reliable method for prediction.

SNF-NN: computational method to predict drug-disease interactions using similarity network fusion and neural networks

BACKGROUND

Drug repositioning is an emerging approach in pharmaceutical research for identifying novel therapeutic potentials for approved drugs and discover therapies for untreated diseases. Due to its time and cost efficiency, drug repositioning plays an instrumental role in optimizing the drug development process compared to the traditional de novo drug discovery process. Advances in the genomics, together with the enormous growth of large-scale publicly available data and the availability of high-performance computing capabilities, have further motivated the development of computational drug repositioning approaches. More recently, the rise of machine learning techniques, together with the availability of powerful computers, has made the area of computational drug repositioning an area of intense activities.

RESULTS

In this study, a novel framework SNF-NN based on deep learning is presented, where novel drug-disease interactions are predicted using drug-related similarity information, disease-related similarity information, and known drug-disease interactions. Heterogeneous similarity information related to drugs and disease is fed to the proposed framework in order to predict novel drug-disease interactions. SNF-NN uses similarity selection, similarity network fusion, and a highly tuned novel neural network model to predict new drug-disease interactions. The robustness of SNF-NN is evaluated by comparing its performance with nine baseline machine learning methods. The proposed framework outperforms all baseline methods ([Formula: see text] = 0.867, and [Formula: see text]=0.876) using stratified 10-fold cross-validation. To further demonstrate the reliability and robustness of SNF-NN, two datasets are used to fairly validate the proposed framework's performance against seven recent state-of-the-art methods for drug-disease interaction prediction. SNF-NN achieves remarkable performance in stratified 10-fold cross-validation with [Formula: see text] ranging from 0.879 to 0.931 and [Formula: see text] from 0.856 to 0.903. Moreover, the efficiency of SNF-NN is verified by validating predicted unknown drug-disease interactions against clinical trials and published studies.

CONCLUSION

In conclusion, computational drug repositioning research can significantly benefit from integrating similarity measures in heterogeneous networks and deep learning models for predicting novel drug-disease interactions. The data and implementation of SNF-NN are available at http://pages.cpsc.ucalgary.ca/ tnjarada/snf-nn.php .

Predicting virus-host association by Kernelized logistic matrix factorization and similarity network fusion

Background

Viruses are closely related to bacteria and human diseases. It is of great significance to predict associations between viruses and hosts for understanding the dynamics and complex functional networks in microbial community. With the rapid development of the metagenomics sequencing, some methods based on sequence similarity and genomic homology have been used to predict associations between viruses and hosts. However, the known virus-host association network was ignored in these methods.

Results

We proposed a kernelized logistic matrix factorization with integrating different information to predict potential virus-host associations on the heterogeneous network (ILMF-VH) which is constructed by connecting a virus network with a host network based on known virus-host associations. The virus network is constructed based on oligonucleotide frequency measurement, and the host network is constructed by integrating oligonucleotide frequency similarity and Gaussian interaction profile kernel similarity through similarity network fusion. The host prediction accuracy of our method is better than other methods. In addition, case studies show that the host of crAssphage predicted by ILMF-VH is consistent with presumed host in previous studies, and another potential host Escherichia coli is also predicted.

Conclusions

The proposed model is an effective computational tool for predicting interactions between viruses and hosts effectively, and it has great potential for discovering novel hosts of viruses.

Similarity Network Fusion: A Novel Application to Making Clinical Diagnoses

Similarity Network Fusion (SNF) is a novel methodological tool that integrates multiple different types of data to identify homogeneous subsets of patients in whom disease classification may be otherwise unclear or challenging. In this review article, the authors hope to provide insight into how SNF can be used in clinical decision making where the aim is to have little influence on the data prior to obtaining the results of the analysis.

Fusing multiple protein-protein similarity networks to effectively predict lncRNA-protein interactions

BACKGROUND

Long non-coding RNA (lncRNA) plays important roles in many biological and pathological processes, including transcriptional regulation and gene regulation. As lncRNA interacts with multiple proteins, predicting lncRNA-protein interactions (lncRPIs) is an important way to study the functions of lncRNA. Up to now, there have been a few works that exploit protein-protein interactions (PPIs) to help the prediction of new lncRPIs.

RESULTS

In this paper, we propose to boost the prediction of lncRPIs by fusing multiple protein-protein similarity networks (PPSNs). Concretely, we first construct four PPSNs based on protein sequences, protein domains, protein GO terms and the STRING database respectively, then build a more informative PPSN by fusing these four constructed PPSNs. Finally, we predict new lncRPIs by a random walk method with the fused PPSN and known lncRPIs. Our experimental results show that the new approach outperforms the existing methods.

CONCLUSION

Fusing multiple protein-protein similarity networks can effectively boost the performance of predicting lncRPIs.

Constructing an integrated gene similarity network for the identification of disease genes

BACKGROUND

Discovering novel genes that are involved human diseases is a challenging task in biomedical research. In recent years, several computational approaches have been proposed to prioritize candidate disease genes. Most of these methods are mainly based on protein-protein interaction (PPI) networks. However, since these PPI networks contain false positives and only cover less half of known human genes, their reliability and coverage are very low. Therefore, it is highly necessary to fuse multiple genomic data to construct a credible gene similarity network and then infer disease genes on the whole genomic scale.

RESULTS

We proposed a novel method, named RWRB, to infer causal genes of interested diseases. First, we construct five individual gene (protein) similarity networks based on multiple genomic data of human genes. Then, an integrated gene similarity network (IGSN) is reconstructed based on similarity network fusion (SNF) method. Finally, we employee the random walk with restart algorithm on the phenotype-gene bilayer network, which combines phenotype similarity network, IGSN as well as phenotype-gene association network, to prioritize candidate disease genes. We investigate the effectiveness of RWRB through leave-one-out cross-validation methods in inferring phenotype-gene relationships. Results show that RWRB is more accurate than state-of-the-art methods on most evaluation metrics. Further analysis shows that the success of RWRB is benefited from IGSN which has a wider coverage and higher reliability comparing with current PPI networks. Moreover, we conduct a comprehensive case study for Alzheimer's disease and predict some novel disease genes that supported by literature.

CONCLUSIONS

RWRB is an effective and reliable algorithm in prioritizing candidate disease genes on the genomic scale. Software and supplementary information are available at http://nclab.hit.edu.cn/~tianzhen/RWRB/ .

Combining Disrupted and Discriminative Topological Properties of Functional Connectivity Networks as Neuroimaging Biomarkers for Accurate Diagnosis of Early Tourette Syndrome Children

Tourette syndrome (TS) is a childhood-onset neurological disorder. To date, accurate TS diagnosis remains challenging due to its varied clinical expressions and dependency on qualitative description of symptoms. Therefore, identifying accurate and objective neuroimaging biomarkers may help improve early TS diagnosis. As resting-state functional MRI (rs-fMRI) has been demonstrated as a promising neuroimaging tool for TS diagnosis, previous rs-fMRI studies on TS revealed functional connectivity (FC) changes in a few local brain networks or circuits. However, no study explored the disrupted topological organization of whole-brain FC networks in TS children. Meanwhile, very few studies have examined brain functional networks using machine-learning methods for diagnostics. In this study, we construct individual whole-brain, ROI-level FC networks for 29 drug-naive TS children and 37 healthy children. Then, we use graph theory analysis to investigate the topological disruptions between groups. The identified disrupted regions in FC networks not only involved the sensorimotor association regions but also the visual, default-mode and language areas, all highly related to TS. Furthermore, we propose a novel classification framework based on similarity network fusion (SNF) algorithm, to both diagnose an individual subject and explore the discriminative power of FC network topological properties in distinguishing between TS children and controls. We achieved a high accuracy of 88.79%, and the involved discriminative regions for classification were also highly related to TS. Together, both the disrupted topological properties between groups and the discriminative topological features for classification may be considered as comprehensive and helpful neuroimaging biomarkers for assisting the clinical TS diagnosis.