PathIntegrate: Multivariate modelling approaches for pathway-based multi-omics data integration.

As terabytes of multi-omics data are being generated, there is an ever-increasing need for methods facilitating the integration and interpretation of such data. Current multi-omics integration methods typically output lists, clusters, or subnetworks of molecules related to an outcome. Even with expert domain knowledge, discerning the biological processes involved is a time-consuming activity. Here we propose PathIntegrate, a method for integrating multi-omics datasets based on pathways, designed to exploit knowledge of biological systems and thus provide interpretable models for such studies. PathIntegrate employs single-sample pathway analysis to transform multi-omics datasets from the molecular to the pathway-level, and applies a predictive single-view or multi-view model to integrate the data. Model outputs include multi-omics pathways ranked by their contribution to the outcome prediction, the contribution of each omics layer, and the importance of each molecule in a pathway. Using semi-synthetic data we demonstrate the benefit of grouping molecules into pathways to detect signals in low signal-to-noise scenarios, as well as the ability of PathIntegrate to precisely identify important pathways at low effect sizes. Finally, using COPD and COVID-19 data we showcase how PathIntegrate enables convenient integration and interpretation of complex high-dimensional multi-omics datasets. PathIntegrate is available as an open-source Python package.

Pathway analysis in metabolomics: Recommendations for the use of over-representation analysis.

Over-representation analysis (ORA) is one of the commonest pathway analysis approaches used for the functional interpretation of metabolomics datasets. Despite the widespread use of ORA in metabolomics, the community lacks guidelines detailing its best-practice use. Many factors have a pronounced impact on the results, but to date their effects have received little systematic attention. Using five publicly available datasets, we demonstrated that changes in parameters such as the background set, differential metabolite selection methods, and pathway database used can result in profoundly different ORA results. The use of a non-assay-specific background set, for example, resulted in large numbers of false-positive pathways. Pathway database choice, evaluated using three of the most popular metabolic pathway databases (KEGG, Reactome, and BioCyc), led to vastly different results in both the number and function of significantly enriched pathways. Factors that are specific to metabolomics data, such as the reliability of compound identification and the chemical bias of different analytical platforms also impacted ORA results. Simulated metabolite misidentification rates as low as 4% resulted in both gain of false-positive pathways and loss of truly significant pathways across all datasets. Our results have several practical implications for ORA users, as well as those using alternative pathway analysis methods. We offer a set of recommendations for the use of ORA in metabolomics, alongside a set of minimal reporting guidelines, as a first step towards the standardisation of pathway analysis in metabolomics.

PathIntegrate: Multivariate modelling approaches for pathway-based multi-omics data integration.

As terabytes of multi-omics data are being generated, there is an ever-increasing need for methods facilitating the integration and interpretation of such data. Current multi-omics integration methods typically output lists, clusters, or subnetworks of molecules related to an outcome. Even with expert domain knowledge, discerning the biological processes involved is a time-consuming activity. Here we propose PathIntegrate, a method for integrating multi-omics datasets based on pathways, designed to exploit knowledge of biological systems and thus provide interpretable models for such studies. PathIntegrate employs single-sample pathway analysis to transform multi-omics datasets from the molecular to the pathway-level, and applies a predictive single-view or multi-view model to integrate the data. Model outputs include multi-omics pathways ranked by their contribution to the outcome prediction, the contribution of each omics layer, and the importance of each molecule in a pathway. Using semi-synthetic data we demonstrate the benefit of grouping molecules into pathways to detect signals in low signal-to-noise scenarios, as well as the ability of PathIntegrate to precisely identify important pathways at low effect sizes. Finally, using COPD and COVID-19 data we showcase how PathIntegrate enables convenient integration and interpretation of complex high-dimensional multi-omics datasets. The PathIntegrate Python package is available at https://github.com/cwieder/PathIntegrate.

Effects of tuberculosis and/or HIV-1 infection on COVID-19 presentation and immune response in Africa.

Few studies from Africa have described the clinical impact of co-infections on SARS-CoV-2 infection. Here, we investigate the presentation and outcome of SARS-CoV-2 infection in an African setting of high HIV-1 and tuberculosis prevalence by an observational case cohort of SARS-CoV-2 patients. A comparator group of non SARS-CoV-2 participants is included. The study includes 104 adults with SARS-CoV-2 infection of whom 29.8% are HIV-1 co-infected. Two or more co-morbidities are present in 57.7% of participants, including HIV-1 (30%) and active tuberculosis (14%). Amongst patients dually infected by tuberculosis and SARS-CoV-2, clinical features can be typical of either SARS-CoV-2 or tuberculosis: lymphopenia is exacerbated, and some markers of inflammation (D-dimer and ferritin) are further elevated (p < 0.05). Amongst HIV-1 co-infected participants those with low CD4 percentage strata exhibit reduced total, but not neutralising, anti-SARS-CoV-2 antibodies. SARS-CoV-2 specific CD8 T cell responses are present in 35.8% participants overall but undetectable in combined HIV-1 and tuberculosis. Death occurred in 30/104 (29%) of all COVID-19 patients and in 6/15 (40%) of patients with coincident SARS-CoV-2 and tuberculosis. This shows that in a high incidence setting, tuberculosis is a common co-morbidity in patients admitted to hospital with COVID-19. The immune response to SARS-CoV-2 is adversely affected by co-existent HIV-1 and tuberculosis.

The In Vivo Transcriptomic Blueprint of Mycobacterium tuberculosis in the Lung.

Mycobacterium tuberculosis (Mtb) genes encoding proteins targeted by vaccines and drugs should be expressed in the lung, the main organ affected by Mtb, for these to be effective. However, the pulmonary expression of most Mtb genes and their proteins remains poorly characterized. The aim of this study is to fill this knowledge gap. We analyzed large scale transcriptomic datasets from specimens of Mtb-infected humans, TB-hypersusceptible (C3H/FeJ) and TB-resistant (C57BL/6J) mice and compared data to in vitro cultured Mtb gene-expression profiles. Results revealed high concordance in the most abundantly in vivo expressed genes between pulmonary Mtb transcriptomes from different datasets and different species. As expected, this contrasted with a lower correlation found with the highest expressed Mtb genes from in vitro datasets. Among the most consistently and highly in vivo expressed genes, 35 have not yet been explored as targets for vaccination or treatment. More than half of these genes are involved in protein synthesis or metabolic pathways. This first lung-oriented multi-study analysis of the in vivo expressed Mtb-transcriptome provides essential data that considerably increase our understanding of pulmonary TB infection biology, and identifies novel molecules for target-based TB-vaccine and drug development.

Standardized approaches for clinical sampling and endpoint ascertainment in tuberculous meningitis studies.

Tuberculous meningitis (TBM), the most severe manifestation of tuberculosis, has poorly understood immunopathology and high mortality and morbidity despite antituberculous therapy. This calls for accelerated clinical and basic science research in this field. As TBM disproportionally affects poorer communities, studies are often performed in resource-limited environments, creating challenges for data collection and harmonisation. Comparison of TBM studies has been hampered by variation in sampling strategies, study design and choice of study endpoints.  Based on literature review and expert consensus, this paper provides firstly, practical recommendations to enable thorough diagnostic, pathophysiological and pharmacokinetic studies using clinical samples, and facilitates better data aggregation and comparisons across populations and settings. Secondly, we discuss clinically relevant study endpoints, including neuroimaging, functional outcome, and cause of death, with suggestions of how these could be applied in different designs for future TBM studies.

Corticosteroids as an adjunct to tuberculosis therapy.

INTRODUCTION: Inflammation, or the prolonged resolution of inflammation, contributes to death from tuberculosis. Interest in inflammatory mechanisms and the prospect of beneficial immune modulation as an adjunct to antibacterial therapy has revived and the concept of host directed therapies has been advanced. Such renewed attention has however, overlooked the experience of such therapy with corticosteroids. Areas covered: The authors conducted literature searches and evaluated randomized clinical trials, systematic reviews and current guidelines and summarize these findings. They found evidence of benefit in meningeal and pericardial tuberculosis in HIV-1 uninfected persons, but less so in those HIV-1 coinfected and evidence of harm in the form of opportunist malignancy in those not prescribed antiretroviral therapy. Adjunctive corticosteroids are however of benefit in the treatment and prevention of paradoxical HIV-tuberculosis immune reconstitution inflammatory syndrome. Expert commentary: Further high-quality clinical trials and experimental medicine studies are warranted and analysis of materials arising from such studies could illuminate ways to improve corticosteroid efficacy or identify novel pathways for more specific intervention.