ADAGE signature analysis: differential expression analysis with data-defined gene sets

MousiPLIER: A Mouse Pathway-Level Information Extractor Model

High-throughput gene expression profiling measures individual gene expression across conditions. However, genes are regulated in complex networks, not as individual entities, limiting the interpretability of gene expression data. Machine learning models that incorporate prior biological knowledge are a powerful tool to extract meaningful biology from gene expression data. Pathway-level information extractor (PLIER) is an unsupervised machine learning method that defines biological pathways by leveraging the vast amount of published transcriptomic data. PLIER converts gene expression data into known pathway gene sets, termed latent variables (LVs), to substantially reduce data dimensionality and improve interpretability. In the current study, we trained the first mouse PLIER model on 190,111 mouse brain RNA-sequencing samples, the greatest amount of training data ever used by PLIER. We then validated the mousiPLIER approach in a study of microglia and astrocyte gene expression across mouse brain aging. mousiPLIER identified biological pathways that are significantly associated with aging, including one latent variable (LV41) corresponding to striatal signal. To gain further insight into the genes contained in LV41, we performed k-means clustering on the training data to identify studies that respond strongly to LV41. We found that the variable was relevant to striatum and aging across the scientific literature. Finally, we built a Web server (http://mousiplier.greenelab.com/) for users to easily explore the learned latent variables. Taken together, this study defines mousiPLIER as a method to uncover meaningful biological processes in mouse brain transcriptomic studies.

Loss of LasR function leads to decreased repression of Pseudomonas aeruginosa PhoB activity at physiological phosphate concentrations

While the Pseudomonas aeruginosa LasR transcription factor plays a role in quorum sensing (QS) across phylogenetically-distinct lineages, isolates with loss-of-function mutations in lasR (LasR- strains) are commonly found in diverse settings including infections where they are associated with worse clinical outcomes. In LasR- strains, the transcription factor RhlR, which is controlled by LasR, can be alternately activated in low inorganic phosphate (Pi) concentrations via the two-component system PhoR-PhoB. Here, we demonstrate a new link between LasR and PhoB in which the absence of LasR increases PhoB activity at physiological Pi concentrations and raises the Pi concentration necessary for PhoB inhibition. PhoB activity was also less repressed by Pi in mutants lacking different QS regulators (RhlR and PqsR) and in mutants lacking genes required for the production of QS-regulated phenazines suggesting that decreased phenazine production was one reason for decreased PhoB repression by Pi in LasR- strains. In addition, the CbrA-CbrB two-component system, which is elevated in LasR- strains, was necessary for reduced PhoB repression by Pi and a Δcrc mutant, which lacks the CbrA-CbrB-controlled translational repressor, activated PhoB at higher Pi concentrations than the wild type. The ΔlasR mutant had a PhoB-dependent growth advantage in a medium with no added Pi and increased virulence-determinant gene expression in a medium with physiological Pi, in part through reactivation of QS. This work suggests PhoB activity may contribute to the virulence of LasR- P. aeruginosa and subsequent clinical outcomes.

Pseudomonas aeruginosa transcriptome analysis of metal restriction in ex vivo cystic fibrosis sputum

Chronic Pseudomonas aeruginosa lung infections are a feature of cystic fibrosis (CF) that many patients experience even with the advent of highly effective modulator therapies. Identifying factors that impact P. aeruginosa in the CF lung could yield novel strategies to eradicate infection or otherwise improve outcomes. To complement published P. aeruginosa studies using laboratory models or RNA isolated from sputum, we analyzed transcripts of strain PAO1 after incubation in sputum from different CF donors prior to RNA extraction. We compared PAO1 gene expression in this "spike-in" sputum model to that for P. aeruginosa grown in synthetic cystic fibrosis sputum medium to determine key genes, which are among the most differentially expressed or most highly expressed. Using the key genes, gene sets with correlated expression were determined using the gene expression analysis tool eADAGE. Gene sets were used to analyze the activity of specific pathways in P. aeruginosa grown in sputum from different individuals. Gene sets that we found to be more active in sputum showed similar activation in published data that included P. aeruginosa RNA isolated from sputum relative to corresponding in vitro reference cultures. In the ex vivo samples, P. aeruginosa had increased levels of genes related to zinc and iron acquisition which were suppressed by metal amendment of sputum. We also found a significant correlation between expression of the H1-type VI secretion system and CFTR corrector use by the sputum donor. An ex vivo sputum model or synthetic sputum medium formulation that imposes metal restriction may enhance future CF-related studies.IMPORTANCEIdentifying the gene expression programs used by Pseudomonas aeruginosa to colonize the lungs of people with cystic fibrosis (CF) will illuminate new therapeutic strategies. To capture these transcriptional programs, we cultured the common P. aeruginosa laboratory strain PAO1 in expectorated sputum from CF patient donors. Through bioinformatic analysis, we defined sets of genes that are more transcriptionally active in real CF sputum compared to a synthetic cystic fibrosis sputum medium. Many of the most differentially active gene sets contained genes related to metal acquisition, suggesting that these gene sets play an active role in scavenging for metals in the CF lung environment which may be inadequately represented in some models. Future studies of P. aeruginosa transcript abundance in CF may benefit from the use of an expectorated sputum model or media supplemented with factors that induce metal restriction.

Meta-Research: Understudied genes are lost in a leaky pipeline between genome-wide assays and reporting of results

Present-day publications on human genes primarily feature genes that already appeared in many publications prior to completion of the Human Genome Project in 2003. These patterns persist despite the subsequent adoption of high-throughput technologies, which routinely identify novel genes associated with biological processes and disease. Although several hypotheses for bias in the selection of genes as research targets have been proposed, their explanatory powers have not yet been compared. Our analysis suggests that understudied genes are systematically abandoned in favor of better-studied genes between the completion of -omics experiments and the reporting of results. Understudied genes remain abandoned by studies that cite these -omics experiments. Conversely, we find that publications on understudied genes may even accrue a greater number of citations. Among 45 biological and experimental factors previously proposed to affect which genes are being studied, we find that 33 are significantly associated with the choice of hit genes presented in titles and abstracts of -omics studies. To promote the investigation of understudied genes, we condense our insights into a tool, find my understudied genes (FMUG), that allows scientists to engage with potential bias during the selection of hits. We demonstrate the utility of FMUG through the identification of genes that remain understudied in vertebrate aging. FMUG is developed in Flutter and is available for download at fmug.amaral.northwestern.edu as a MacOS/Windows app.

Meta-Research: understudied genes are lost in a leaky pipeline between genome-wide assays and reporting of results

Present-day publications on human genes primarily feature genes that already appeared in many publications prior to completion of the Human Genome Project in 2003. These patterns persist despite the subsequent adoption of high-throughput technologies, which routinely identify novel genes associated with biological processes and disease. Although several hypotheses for bias in the selection of genes as research targets have been proposed, their explanatory powers have not yet been compared. Our analysis suggests that understudied genes are systematically abandoned in favor of better-studied genes between the completion of -omics experiments and the reporting of results. Understudied genes remain abandoned by studies that cite these -omics experiments. Conversely, we find that publications on understudied genes may even accrue a greater number of citations. Among 45 biological and experimental factors previously proposed to affect which genes are being studied, we find that 33 are significantly associated with the choice of hit genes presented in titles and abstracts of - omics studies. To promote the investigation of understudied genes we condense our insights into a tool, find my understudied genes (FMUG), that allows scientists to engage with potential bias during the selection of hits. We demonstrate the utility of FMUG through the identification of genes that remain understudied in vertebrate aging. FMUG is developed in Flutter and is available for download at fmug.amaral.northwestern.edu as a MacOS/Windows app.

Analysis of Pseudomonas aeruginosa transcription in an ex vivo cystic fibrosis sputum model identifies metal restriction as a gene expression stimulus

Chronic Pseudomonas aeruginosa lung infections are a distinctive feature of cystic fibrosis (CF) pathology, that challenge adults with CF even with the advent of highly effective modulator therapies. Characterizing P. aeruginosa transcription in the CF lung and identifying factors that drive gene expression could yield novel strategies to eradicate infection or otherwise improve outcomes. To complement published P. aeruginosa gene expression studies in laboratory culture models designed to model the CF lung environment, we employed an ex vivo sputum model in which laboratory strain PAO1 was incubated in sputum from different CF donors. As part of the analysis, we compared PAO1 gene expression in this "spike-in" sputum model to that for P. aeruginosa grown in artificial sputum medium (ASM). Analyses focused on genes that were differentially expressed between sputum and ASM and genes that were most highly expressed in sputum. We present a new approach that used sets of genes with correlated expression, identified by the gene expression analysis tool eADAGE, to analyze the differential activity of pathways in P. aeruginosa grown in CF sputum from different individuals. A key characteristic of P. aeruginosa grown in expectorated CF sputum was related to zinc and iron acquisition, but this signal varied by donor sputum. In addition, a significant correlation between P. aeruginosa expression of the H1-type VI secretion system and corrector use by the sputum donor was observed. These methods may be broadly useful in looking for variable signals across clinical samples.

Computationally Efficient Assembly of Pseudomonas aeruginosa Gene Expression Compendia

Thousands of Pseudomonas aeruginosa RNA sequencing (RNA-seq) gene expression profiles are publicly available via the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA). In this work, the transcriptional profiles from hundreds of studies performed by over 75 research groups were reanalyzed in aggregate to create a powerful tool for hypothesis generation and testing. Raw sequence data were uniformly processed using the Salmon pseudoaligner, and this read mapping method was validated by comparison to a direct alignment method. We developed filtering criteria to exclude samples with aberrant levels of housekeeping gene expression or an unexpected number of genes with no reported values and normalized the filtered compendia using the ratio-of-medians method. The filtering and normalization steps greatly improved gene expression correlations for genes within the same operon or regulon across the 2,333 samples. Since the RNA-seq data were generated using diverse strains, we report the effects of mapping samples to noncognate reference genomes by separately analyzing all samples mapped to cDNA reference genomes for strains PAO1 and PA14, two divergent strains that were used to generate most of the samples. Finally, we developed an algorithm to incorporate new data as they are deposited into the SRA. Our processing and quality control methods provide a scalable framework for taking advantage of the troves of biological information hibernating in the depths of microbial gene expression data and yield useful tools for P. aeruginosa RNA-seq data to be leveraged for diverse research goals. IMPORTANCE Pseudomonas aeruginosa is a causative agent of a wide range of infections, including chronic infections associated with cystic fibrosis. These P. aeruginosa infections are difficult to treat and often have negative outcomes. To aid in the study of this problematic pathogen, we mapped, filtered for quality, and normalized thousands of P. aeruginosa RNA-seq gene expression profiles that were publicly available via the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA). The resulting compendia facilitate analyses across experiments, strains, and conditions. Ultimately, the workflow that we present could be applied to analyses of other microbial species.

Using genome-wide expression compendia to study microorganisms

A gene expression compendium is a heterogeneous collection of gene expression experiments assembled from data collected for diverse purposes. The widely varied experimental conditions and genetic backgrounds across samples creates a tremendous opportunity for gaining a systems level understanding of the transcriptional responses that influence phenotypes. Variety in experimental design is particularly important for studying microbes, where the transcriptional responses integrate many signals and demonstrate plasticity across strains including response to what nutrients are available and what microbes are present. Advances in high-throughput measurement technology have made it feasible to construct compendia for many microbes. In this review we discuss how these compendia are constructed and analyzed to reveal transcriptional patterns.

Review of Artificial Intelligence Applications and Algorithms for Brain Organoid Research

The human brain organoid is a miniature three-dimensional tissue culture that can simulate the structure and function of the brain in an in vitro culture environment. Although we consider that human brain organoids could be used to understand brain development and diseases, experimental models of human brain organoids are so highly variable that we apply artificial intelligence (AI) techniques to investigate the development mechanism of the human brain. Therefore, this study briefly reviewed commonly used AI applications for human brain organoid-magnetic resonance imaging, electroencephalography, and gene editing techniques, as well as related AI algorithms. Finally, we discussed the limitations, challenges, and future study direction of AI-based technology for human brain organoids.

Conditional antagonism in co-cultures of Pseudomonas aeruginosa and Candida albicans: An intersection of ethanol and phosphate signaling distilled from dual-seq transcriptomics

Pseudomonas aeruginosa and Candida albicans are opportunistic pathogens whose interactions involve the secreted products ethanol and phenazines. Here, we describe the role of ethanol in mixed-species co-cultures by dual-seq analyses. P. aeruginosa and C. albicans transcriptomes were assessed after growth in mono-culture or co-culture with either ethanol-producing C. albicans or a C. albicans mutant lacking the primary ethanol dehydrogenase, Adh1. Analysis of the RNA-Seq data using KEGG pathway enrichment and eADAGE methods revealed several P. aeruginosa responses to C. albicans-produced ethanol including the induction of a non-canonical low-phosphate response regulated by PhoB. C. albicans wild type, but not C. albicans adh1Δ/Δ, induces P. aeruginosa production of 5-methyl-phenazine-1-carboxylic acid (5-MPCA), which forms a red derivative within fungal cells and exhibits antifungal activity. Here, we show that C. albicans adh1Δ/Δ no longer activates P. aeruginosa PhoB and PhoB-regulated phosphatase activity, that exogenous ethanol complements this defect, and that ethanol is sufficient to activate PhoB in single-species P. aeruginosa cultures at permissive phosphate levels. The intersection of ethanol and phosphate in co-culture is inversely reflected in C. albicans; C. albicans adh1Δ/Δ had increased expression of genes regulated by Pho4, the C. albicans transcription factor that responds to low phosphate, and Pho4-dependent phosphatase activity. Together, these results show that C. albicans-produced ethanol stimulates P. aeruginosa PhoB activity and 5-MPCA-mediated antagonism, and that both responses are dependent on local phosphate concentrations. Further, our data suggest that phosphate scavenging by one species improves phosphate access for the other, thus highlighting the complex dynamics at play in microbial communities.

Pseudomonas aeruginosa and Candida albicans are opportunistic pathogens that are frequently isolated from co-infections. Using a combination of dual-seq transcriptomics and genetics approaches, we found that ethanol produced by C. albicans stimulates the PhoB regulon in P. aeruginosa asynchronously with activation of the Pho4 regulon in C. albicans. We validated our result by showing that PhoB plays multiple roles in co-culture including orchestrating the competition for phosphate and the production of 5-methyl-phenazine-1-carboxylic acid; the P. aeruginosa phenazine response to C. albicans-produced ethanol depends on phosphate availability. The conditional stimulation of antifungal production in response to sub-inhibitory concentrations of ethanol only under phosphate limitation highlights the importance of considering nutrient concentrations in the analysis of co-culture interactions and suggests that the low-phosphate response in one species influences phosphate availability for the other.

Integrative Analysis Identifies Candidate Tumor Microenvironment and Intracellular Signaling Pathways that Define Tumor Heterogeneity in NF1

Neurofibromatosis type 1 (NF1) is a monogenic syndrome that gives rise to numerous symptoms including cognitive impairment, skeletal abnormalities, and growth of benign nerve sheath tumors. Nearly all NF1 patients develop cutaneous neurofibromas (cNFs), which occur on the skin surface, whereas 40-60% of patients develop plexiform neurofibromas (pNFs), which are deeply embedded in the peripheral nerves. Patients with pNFs have a ~10% lifetime chance of these tumors becoming malignant peripheral nerve sheath tumors (MPNSTs). These tumors have a severe prognosis and few treatment options other than surgery. Given the lack of therapeutic options available to patients with these tumors, identification of druggable pathways or other key molecular features could aid ongoing therapeutic discovery studies. In this work, we used statistical and machine learning methods to analyze 77 NF1 tumors with genomic data to characterize key signaling pathways that distinguish these tumors and identify candidates for drug development. We identified subsets of latent gene expression variables that may be important in the identification and etiology of cNFs, pNFs, other neurofibromas, and MPNSTs. Furthermore, we characterized the association between these latent variables and genetic variants, immune deconvolution predictions, and protein activity predictions.

Loss of RNA Chaperone Hfq Unveils a Toxic Pathway in Pseudomonas aeruginosa

Hfq is an RNA chaperone that serves as a master regulator of bacterial physiology. Here we show that in the opportunistic pathogen Pseudomonas aeruginosa, the loss of Hfq can result in a dramatic reduction in growth in a manner that is dependent upon MexT, a transcription regulator that governs antibiotic resistance in this organism. Using a combination of chromatin immunoprecipitation with high-throughput sequencing and transposon insertion sequencing, we identify the MexT-activated genes responsible for mediating the growth defect of hfq mutant cells. These include a newly identified MexT-controlled gene that we call hilR We demonstrate that hilR encodes a small protein that is acutely toxic to wild-type cells when produced ectopically. Furthermore, we show that hilR expression is negatively regulated by Hfq, offering a possible explanation for the growth defect of hfq mutant cells. Finally, we present evidence that the expression of MexT-activated genes is dependent upon GshA, an enzyme involved in the synthesis of glutathione. Our findings suggest that Hfq can influence the growth of P. aeruginosa by limiting the toxic effects of specific MexT-regulated genes. Moreover, our results identify glutathione to be a factor important for the in vivo activity of MexT.IMPORTANCE Here we show that the conserved RNA chaperone Hfq is important for the growth of the opportunistic pathogen Pseudomonas aeruginosa We found that the growth defect of hfq mutant cells is dependent upon the expression of genes that are under the control of the transcription regulator MexT. These include a gene that we refer to as hilR, which we show is negatively regulated by Hfq and encodes a small protein that can be toxic when ectopically produced in wild-type cells. Thus, Hfq can influence the growth of P. aeruginosa by limiting the toxic effects of MexT-regulated genes, including one encoding a previously unrecognized small protein. We also show that MexT activity depends on an enzyme that synthesizes glutathione.

Ethanol Stimulates Trehalose Production through a SpoT-DksA-AlgU-Dependent Pathway in Pseudomonas aeruginosa

Pseudomonas aeruginosa frequently resides among ethanol-producing microbes, making its response to the microbially produced concentrations of ethanol relevant to understanding its biology. Our transcriptome analysis found that genes involved in trehalose metabolism were induced by low concentrations of ethanol, and biochemical assays showed that levels of intracellular trehalose increased significantly upon growth with ethanol. The increase in trehalose was dependent on the TreYZ pathway but not other trehalose-metabolic enzymes (TreS or TreA). The sigma factor AlgU (AlgT), a homolog of RpoE in other species, was required for increased expression of the treZ gene and trehalose levels, but induction was not controlled by the well-characterized proteolysis of its anti-sigma factor, MucA. Growth with ethanol led to increased SpoT-dependent (p)ppGpp accumulation, which stimulates AlgU-dependent transcription of treZ and other AlgU-regulated genes through DksA, a (p)ppGpp and RNA polymerase binding protein. Ethanol stimulation of trehalose also required acylhomoserine lactone (AHL)-mediated quorum sensing (QS), as induction was not observed in a ΔlasR ΔrhlR strain. A network analysis using a model, eADAGE, built from publicly available P. aeruginosa transcriptome data sets (J. Tan, G. Doing, K. A. Lewis, C. E. Price, et al., Cell Syst 5:63-71, 2017, https://doi.org/10.1016/j.cels.2017.06.003) provided strong support for our model in which treZ and coregulated genes are controlled by both AlgU- and AHL-mediated QS. Consistent with (p)ppGpp- and AHL-mediated quorum-sensing regulation, ethanol, even when added at the time of culture inoculation, stimulated treZ transcript levels and trehalose production in cells from post-exponential-phase cultures but not in cells from exponential-phase cultures. These data highlight the integration of growth and cell density cues in the P. aeruginosa transcriptional response to ethanol.IMPORTANCE Pseudomonas aeruginosa is often found with bacteria and fungi that produce fermentation products, including ethanol. At concentrations similar to those produced by environmental microbes, we found that ethanol stimulated expression of trehalose-biosynthetic genes and cellular levels of trehalose, a disaccharide that protects against environmental stresses. The induction of trehalose by ethanol required the alternative sigma factor AlgU through DksA- and SpoT-dependent (p)ppGpp. Trehalose accumulation also required AHL quorum sensing and occurred only in post-exponential-phase cultures. This work highlights how cells integrate cell density and growth cues in their responses to products made by other microbes and reveals a new role for (p)ppGpp in the regulation of AlgU activity.