Frontiers in systems biology最新文献

Post-Acute Sequelae of SARS-CoV-2 infection (PASC or "Long COVID"), includes numerous chronic conditions associated with widespread morbidity and rising healthcare costs. PASC has highly variable clinical presentations, and likely includes multiple molecular subtypes, but it remains poorly understood from a molecular and mechanistic standpoint. This hampers the development of rationally targeted therapeutic strategies. The NIH-sponsored "Researching COVID to Enhance Recovery" (RECOVER) initiative includes several retrospective/prospective observational cohort studies enrolling adult, pregnant adult and pediatric patients respectively. RECOVER formed an "OMICS" multidisciplinary task force, including clinicians, pathologists, laboratory scientists and data scientists, charged with developing recommendations to apply cutting-edge system biology technologies to achieve the goals of RECOVER. The task force met biweekly over 14 months, to evaluate published evidence, examine the possible contribution of each "omics" technique to the study of PASC and develop study design recommendations. The OMICS task force recommended an integrated, longitudinal, simultaneous systems biology study of participant biospecimens on the entire RECOVER cohorts through centralized laboratories, as opposed to multiple smaller studies using one or few analytical techniques. The resulting multi-dimensional molecular dataset should be correlated with the deep clinical phenotyping performed through RECOVER, as well as with information on demographics, comorbidities, social determinants of health, the exposome and lifestyle factors that may contribute to the clinical presentations of PASC. This approach will minimize lab-to-lab technical variability, maximize sample size for class discovery, and enable the incorporation of as many relevant variables as possible into statistical models. Many of our recommendations have already been considered by the NIH through the peer-review process, resulting in the creation of a systems biology panel that is currently designing the studies we proposed. This system biology strategy, coupled with modern data science approaches, will dramatically improve our prospects for accurate disease subtype identification, biomarker discovery and therapeutic target identification for precision treatment. The resulting dataset should be made available to the scientific community for secondary analyses. Analogous system biology approaches should be built into the study designs of large observational studies whenever possible.

Introduction: The availability of large-scale multi-omic data has revolution-ized the study of cellular machinery, enabling a systematic understanding of biological processes. However, the integration of these datasets into Genome-Scale Models of Metabolism (GEMs) re-mains underexplored. Existing methods often link transcriptome and proteome data independently to reaction boundaries, providing models with estimated maximum reaction rates based on individual datasets. This independent approach, however, introduces uncertainties and inaccuracies.

Methods: To address these challenges, we applied a principal component analysis (PCA)-based approach to integrate transcriptome and proteome data. This method facilitates the reconstruction of context-specific models grounded in multi-omics data, enhancing their biological relevance and predictive capacity.

Results: Using this approach, we successfully reconstructed an astrocyte GEM with improved prediction capabilities compared to state-of-the-art models available in the literature.

Discussion: These advancements underscore the potential of multi-omic inte-gration to refine metabolic modeling and its critical role in studying neurodegeneration and developing effective therapies.

Gaussian Graphical Models (GGMs) are a type of network modeling that uses partial correlation rather than correlation for representing complex relationships among multiple variables. The advantage of using partial correlation is to show the relation between two variables after "adjusting" for the effects of other variables and leads to more parsimonious and interpretable models. There are well established procedures to build GGMs from a sample of independent and identical distributed observations. However, many studies include clustered and longitudinal data that result in correlated observations and ignoring this correlation among observations can lead to inflated Type I error. In this paper, we propose a cluster-based bootstrap algorithm to infer GGMs from correlated data. We use extensive simulations of correlated data from family-based studies to show that the proposed bootstrap method does not inflate the Type I error while retaining statistical power compared to alternative solutions when there are sufficient number of clusters. We apply our method to learn the GGM that represents complex relations between 47 Polygenic Risk Scores generated using genome-wide genotype data from the Long Life Family Study. By comparing it to the conventional methods that ignore within-cluster correlation, we show that our method controls the Type I error well without power loss.

Immune-mediated diseases are characterized by aberrant immune responses, posing significant challenges to global health. In both inflammatory and autoimmune diseases, dysregulated immune reactions mediated by tissue-residing immune and non-immune cells precipitate chronic inflammation and tissue damage that is amplified by peripheral immune cell extravasation into the tissue. Chemokine receptors are pivotal in orchestrating immune cell migration, yet deciphering the signaling code across cell types, diseases and tissues remains an open challenge. To delineate disease-specific cell-cell communications involved in immune cell migration, we conducted a meta-analysis of publicly available single-cell RNA sequencing (scRNA-seq) data across diverse immune diseases and tissues. Our comprehensive analysis spanned multiple immune disorders affecting major organs: atopic dermatitis and psoriasis (skin), chronic obstructive pulmonary disease and idiopathic pulmonary fibrosis (lung), ulcerative colitis (colon), IgA nephropathy and lupus nephritis (kidney). By interrogating ligand-receptor (L-R) interactions, alterations in cell proportions, and differential gene expression, we unveiled disease-specific and common cell-cell communications involved in chemotaxis and extravasation to shed light on shared immune responses across tissues and diseases. Further, we performed experimental validation of two understudied cell-cell communications. Insights gleaned from this meta-analysis hold promise for the development of targeted therapeutics aimed at modulating immune cell migration to mitigate inflammation and tissue damage. This nuanced understanding of immune cell dynamics at the single-cell resolution opens avenues for precision medicine in immune disease management.

As a driving force of the fourth industrial revolution, deep neural networks are now widely used in various areas of science and technology. Despite the success of deep neural networks in making accurate predictions, their interpretability remains a mystery to researchers. From a statistical point of view, how to conduct statistical inference (e.g., hypothesis testing) based on deep neural networks is still unknown. In this paper, goodness-of-fit statistics are proposed based on commonly used ReLU neural networks, and their potential to test significant input features is explored. A simulation study demonstrates that the proposed test statistic has higher power compared to the commonly used t-test in linear regression when the underlying signal is nonlinear, while controlling the type I error at the desired level. The testing procedure is also applied to gene expression data from the Alzheimer's Disease Neuroimaging Initiative (ADNI).

Circular RNAs (circRNAs) have gained prominence as important players in various biological processes such as gastric cancer (GC). Identification of several dysregulated circRNAs may serve as biomarkers for early diagnosis or as novel therapeutic targets. Predictive models can suggest potential new interactions and regulatory roles of circRNAs in GCs. Experimental validations of key interactions are being performed using in vitro models, confirming the significance of identified circRNA networks. The aim of this review is to highlight the important circRNAs associated with GC. On top of that an overview of the mechanistic details of the biogenesis and functionalities of the circRNAs are also presented. Furthermore, the potentialities of the circRNAs in the field of new drug discovery are deciphered.

Introduction: Pediatric Acute Liver Failure (PALF) presents as a rapidly evolving, multifaceted, and devastating clinical syndrome whose precise etiology remains incompletely understood. Consequently, predicting outcomes-whether survival or mortality-and informing liver transplantation decisions in PALF remain challenging. We have previously implicated High-Mobility Group Box 1 (HMGB1) as a central mediator in PALF-associated dynamic inflammation networks that could be recapitulated in acetaminophen (APAP)-treated mouse hepatocytes (HC) in vitro. Here, we hypothesized that Growth/Differentiation Factor-15 (GDF-15) is involved along with HMGB1 in PALF.

Methods: 28 and 23 inflammatory mediators including HMGB1 and GDF15 were measured in serum samples from PALF patients and cell supernatants from wild-type (C57BL/6) mouse hepatocytes (HC) and from cells from HC-specific HMGB1-null mice (HC-HMGB1^-/-) exposed to APAP, respectively. Results were analyzed computationally to define statistically significant and potential causal relationships.

Results: Circulating GDF-15 was elevated significantly (P < 0.05) in PALF non-survivors as compared to survivors, and together with HMGB1 was identified as a central node in dynamic inflammatory networks in both PALF patients and mouse HC. This analysis also pointed to MIG/CXCL9 as a differential node linking HMGB1 and GDF-15 in survivors but not in non-survivors, and, when combined with in vitro studies, suggested that MIG suppresses GDF-15-induced inflammation.

Discussion: This study suggests GDF-15 as a novel PALF outcome biomarker, posits GDF-15 alongside HMGB1 as a central node within the intricate web of systemic inflammation dynamics in PALF, and infers a novel, negative regulatory role for MIG.

Uncertainty is ubiquitous in biological systems. For example, since gene expression is intrinsically governed by noise, nature shows a fascinating degree of variability. If we want to use a model to predict the behaviour of such an intrinsically stochastic system, we have to cope with the fact that the model parameters are never exactly known, but vary according to some distribution. A key question is then to determine how the uncertainties in the parameters affect the model outcome. Knowing the latter uncertainties is crucial when a model is used for, e.g., experimental design, optimisation, or decision-making. To establish how parameter and model prediction uncertainties are related, Monte Carlo approaches could be used. Then, the model is evaluated for a huge number of parameters sets, drawn from the multivariate parameter distribution. However, when model solutions are computationally expensive this approach is intractable. To overcome this problem, so-called spectral expansion (SE) methods have been developed to quantify prediction uncertainty within a probabilistic framework. Such SE methods have a basis in, e.g., computational mathematics, engineering, physics, and fluid dynamics, and, to a lesser extent, systems biology. The computational costs of SE schemes mainly stem from the calculation of the expansion coefficients. Furthermore, SE effectively leads to a surrogate model which captures the dependence of the model on the uncertainty parameters, but is much simpler to execute compared to the original model. In this paper, we present an innovative scheme for the calculation of the expansion coefficients. It guarantees that the model has to be evaluated only a restricted number of times. Especially for models of high complexity this may be a huge computational advantage. By applying the scheme to a variety of examples we show its power, especially in challenging situations where solutions slowly converge due to high computational costs, bifurcations, and discontinuities.

In the context of in silico clinical trials, mechanistic computer models for pathophysiology and pharmacology (here Quantitative Systems Pharmacology models, QSP) can greatly support the decision making for drug candidates and elucidate the (potential) response of patients to existing and novel treatments. These models are built on disease mechanisms and then parametrized using (clinical study) data. Clinical variability among patients is represented by alternative model parameterizations, called virtual patients. Despite the complexity of disease modeling itself, using individual patient data to build these virtual patients is particularly challenging given the high-dimensional, potentially sparse and noisy clinical trial data. In this work, we investigate the applicability of simulation-based inference (SBI), an advanced probabilistic machine learning approach, for virtual patient generation from individual patient data and we develop and evaluate the concept of nearest patient fits (SBI NPF), which further enhances the fitting performance. At the example of rheumatoid arthritis where prediction of treatment response is notoriously difficult, our experiments demonstrate that the SBI approaches can capture large inter-patient variability in clinical data and can compete with standard fitting methods in the field. Moreover, since SBI learns a probability distribution over the virtual patient parametrization, it naturally provides the probability for alternative parametrizations. The learned distributions allow us to generate highly probable alternative virtual patient populations for rheumatoid arthritis, which could potentially enhance the assessment of drug candidates if used for in silico trials.