noisyR: enhancing biological signal in sequencing datasets by characterizing random technical noise

Mitochondrial complex I activity in microglia sustains neuroinflammation

Sustained smouldering, or low-grade activation, of myeloid cells is a common hallmark of several chronic neurological diseases, including multiple sclerosis1. Distinct metabolic and mitochondrial features guide the activation and the diverse functional states of myeloid cells2. However, how these metabolic features act to perpetuate inflammation of the central nervous system is unclear. Here, using a multiomics approach, we identify a molecular signature that sustains the activation of microglia through mitochondrial complex I activity driving reverse electron transport and the production of reactive oxygen species. Mechanistically, blocking complex I in pro-inflammatory microglia protects the central nervous system against neurotoxic damage and improves functional outcomes in an animal disease model in vivo. Complex I activity in microglia is a potential therapeutic target to foster neuroprotection in chronic inflammatory disorders of the central nervous system3.

Proinflammatory cytokines driving cardiotoxicity in COVID-19

AIMS

Cardiac involvement is common in patients hospitalized with COVID-19 and correlates with an adverse disease trajectory. While cardiac injury has been attributed to direct viral cytotoxicity, serum-induced cardiotoxicity secondary to serological hyperinflammation constitutes a potentially amenable mechanism that remains largely unexplored.

METHODS AND RESULTS

To investigate serological drivers of cardiotoxicity in COVID-19 we have established a robust bioassay that assessed the effects of serum from COVID-19 confirmed patients on human embryonic stem cell (hESC)-derived cardiomyocytes. We demonstrate that serum from COVID-19 positive patients significantly reduced cardiomyocyte viability independent of viral transduction, an effect that was also seen in non-COVID-19 acute respiratory distress syndrome (ARDS). Serum from patients with greater disease severity led to worse cardiomyocyte viability and this significantly correlated with levels of key inflammatory cytokines, including IL-6, TNF-α, IL1-β, IL-10, CRP, and neutrophil to lymphocyte ratio with a specific reduction of CD4+ and CD8+ cells. Combinatorial blockade of IL-6 and TNF-α partly rescued the phenotype and preserved cardiomyocyte viability and function. Bulk RNA sequencing of serum-treated cardiomyocytes elucidated specific pathways involved in the COVID-19 response impacting cardiomyocyte viability, structure, and function. The observed effects of serum-induced cytotoxicity were cell-type selective as serum exposure did not adversely affect microvascular endothelial cell viability but resulted in endothelial activation and a procoagulant state.

CONCLUSION

These results provide direct evidence that inflammatory cytokines are at least in part responsible for the cardiovascular damage seen in COVID-19 and characterise the downstream activated pathways in human cardiomyocytes. The serum signature of patients with severe disease indicates possible targets for therapeutic intervention.

Drug target prediction through deep learning functional representation of gene signatures

Many machine learning applications in bioinformatics currently rely on matching gene identities when analyzing input gene signatures and fail to take advantage of preexisting knowledge about gene functions. To further enable comparative analysis of OMICS datasets, including target deconvolution and mechanism of action studies, we develop an approach that represents gene signatures projected onto their biological functions, instead of their identities, similar to how the word2vec technique works in natural language processing. We develop the Functional Representation of Gene Signatures (FRoGS) approach by training a deep learning model and demonstrate that its application to the Broad Institute's L1000 datasets results in more effective compound-target predictions than models based on gene identities alone. By integrating additional pharmacological activity data sources, FRoGS significantly increases the number of high-quality compound-target predictions relative to existing approaches, many of which are supported by in silico and/or experimental evidence. These results underscore the general utility of FRoGS in machine learning-based bioinformatics applications. Prediction networks pre-equipped with the knowledge of gene functions may help uncover new relationships among gene signatures acquired by large-scale OMICs studies on compounds, cell types, disease models, and patient cohorts.

Human pancreatic islet microRNAs implicated in diabetes and related traits by large-scale genetic analysis

Genetic studies have identified ≥240 loci associated with the risk of type 2 diabetes (T2D), yet most of these loci lie in non-coding regions, masking the underlying molecular mechanisms. Recent studies investigating mRNA expression in human pancreatic islets have yielded important insights into the molecular drivers of normal islet function and T2D pathophysiology. However, similar studies investigating microRNA (miRNA) expression remain limited. Here, we present data from 63 individuals, the largest sequencing-based analysis of miRNA expression in human islets to date. We characterized the genetic regulation of miRNA expression by decomposing the expression of highly heritable miRNAs into cis- and trans-acting genetic components and mapping cis-acting loci associated with miRNA expression [miRNA-expression quantitative trait loci (eQTLs)]. We found i) 84 heritable miRNAs, primarily regulated by trans-acting genetic effects, and ii) 5 miRNA-eQTLs. We also used several different strategies to identify T2D-associated miRNAs. First, we colocalized miRNA-eQTLs with genetic loci associated with T2D and multiple glycemic traits, identifying one miRNA, miR-1908, that shares genetic signals for blood glucose and glycated hemoglobin (HbA1c). Next, we intersected miRNA seed regions and predicted target sites with credible set SNPs associated with T2D and glycemic traits and found 32 miRNAs that may have altered binding and function due to disrupted seed regions. Finally, we performed differential expression analysis and identified 14 miRNAs associated with T2D status-including miR-187-3p, miR-21-5p, miR-668, and miR-199b-5p-and 4 miRNAs associated with a polygenic score for HbA1c levels-miR-216a, miR-25, miR-30a-3p, and miR-30a-5p.

bulkAnalyseR: an accessible, interactive pipeline for analysing and sharing bulk multi-modal sequencing data

Bulk sequencing experiments (single- and multi-omics) are essential for exploring wide-ranging biological questions. To facilitate interactive, exploratory tasks, coupled with the sharing of easily accessible information, we present bulkAnalyseR, a package integrating state-of-the-art approaches using an expression matrix as the starting point (pre-processing functions are available as part of the package). Static summary images are replaced with interactive panels illustrating quality-checking, differential expression analysis (with noise detection) and biological interpretation (enrichment analyses, identification of expression patterns, followed by inference and comparison of regulatory interactions). bulkAnalyseR can handle different modalities, facilitating robust integration and comparison of cis-, trans- and customised regulatory networks.

Myeloid cell interferon secretion restricts Zika flavivirus infection of developing and malignant human neural progenitor cells

Zika virus (ZIKV) can infect human developing brain (HDB) progenitors resulting in epidemic microcephaly, whereas analogous cellular tropism offers treatment potential for the adult brain cancer, glioblastoma (GBM). We compared productive ZIKV infection in HDB and GBM primary tissue explants that both contain SOX2+ neural progenitors. Strikingly, although the HDB proved uniformly vulnerable to ZIKV infection, GBM was more refractory, and this correlated with an innate immune expression signature. Indeed, GBM-derived CD11b+ microglia/macrophages were necessary and sufficient to protect progenitors against ZIKV infection in a non-cell autonomous manner. Using SOX2+ GBM cell lines, we found that CD11b+-conditioned medium containing type 1 interferon beta (IFNβ) promoted progenitor resistance to ZIKV, whereas inhibition of JAK1/2 signaling restored productive infection. Additionally, CD11b+ conditioned medium, and IFNβ treatment rendered HDB progenitor lines and explants refractory to ZIKV. These findings provide insight into neuroprotection for HDB progenitors as well as enhanced GBM oncolytic therapies.

The Sum of Two Halves May Be Different from the Whole-Effects of Splitting Sequencing Samples Across Lanes

The advances in high-throughput sequencing (HTS) have enabled the characterisation of biological processes at an unprecedented level of detail; most hypotheses in molecular biology rely on analyses of HTS data. However, achieving increased robustness and reproducibility of results remains a main challenge. Although variability in results may be introduced at various stages, e.g., alignment, summarisation or detection of differential expression, one source of variability was systematically omitted: the sequencing design, which propagates through analyses and may introduce an additional layer of technical variation. We illustrate qualitative and quantitative differences arising from splitting samples across lanes on bulk and single-cell sequencing. For bulk mRNAseq data, we focus on differential expression and enrichment analyses; for bulk ChIPseq data, we investigate the effect on peak calling and the peaks' properties. At the single-cell level, we concentrate on identifying cell subpopulations. We rely on markers used for assigning cell identities; both smartSeq and 10× data are presented. The observed reduction in the number of unique sequenced fragments limits the level of detail on which the different prediction approaches depend. Furthermore, the sequencing stochasticity adds in a weighting bias corroborated with variable sequencing depths and (yet unexplained) sequencing bias. Subsequently, we observe an overall reduction in sequencing complexity and a distortion in the biological signal across technologies, experimental contexts, organisms and tissues.

Host respiratory transcriptome signature associated with poor outcome in children with influenza-Staphylococcus aureus pneumonia

Respiratory coinfection of influenza with Staphylococcus aureus often causes severe disease; methicillin resistant S. aureus (MRSA) coinfection is frequently fatal. Understanding disease pathogenesis may inform therapies. We aimed to identify host and pathogen transcriptomic (mRNA) signatures from the respiratory compartment of patients with influenza-S. aureus coinfection (ISAC) critical illness that predict worse outcomes. mRNA extracted from endotracheal aspirates was evaluated for S. aureus and host transcriptomic biosignatures. Influenza-MRSA outcomes were worse, but of 190 S. aureus virulence-associated genes, 6 were differentially expressed between MRSA- versus methicillin-susceptible S. aureus coinfected patients and none discriminated outcome. Host gene expression in ISAC patients was compared to influenza infection alone. Patients with poor clinical outcomes (death or prolonged multi-organ dysfunction) had relatively reduced expression of interferons and down-regulation of interferon gamma-induced immune cell chemoattractants CXCL10 and CXCL11. In influenza-S. aureus respiratory coinfection, airway host but not pathogen gene expression profiles predicted worse clinical outcomes.