NPJ Systems Biology and Applications最新文献

Petri nets are commonly applied in modeling biological systems. However, construction of a Petri net model for complex biological systems is often time consuming, and requires expertise in the research area, limiting their application. To address this challenge, we developed GINtoSPN, an R package that automates the conversion of multi-omics molecular interaction network extracted from the Global Integrative Network (GIN) into Petri nets in GraphML format. These GraphML files can be directly used for Signaling Petri Net (SPN) simulation. To demonstrate the utility of this tool, we built a Petri net model for neurofibromatosis type I. Simulation of NF1 gene knockout, compared to normal skin fibroblast cells, revealed persistent accumulation of Ras-GTPs as expected. Additionally, we identified several other genes substantially affected by the loss of NF1's function, exhibiting individual-specific variability. These results highlight the effectiveness of GINtoSPN in streamlining the modeling and simulation of complex biological systems.

Since nail psoriasis restricts the patient's daily activities, therapeutic intervention based on reliable and reproducible evaluation is critical. The Nail Psoriasis Severity Index (NAPSI) is a validated scoring tool, but its usefulness is limited by interobserver variability. This study aimed to develop a reliable and accurate NAPSI scoring tool using deep learning. The tool "NAPSI calculator" includes two parts: nail detection from images and NAPSI scoring. NAPSI was annotated by nine nail experts who are board-certified dermatologists with sufficient experience in a specialized clinic for nail diseases. In the final test set, the "NAPSI calculator" correctly located 137/138 nails and scored NAPSI with higher accuracy than the compared six non-board-certified residents: 83.9% vs 65.7%; P = 0.008 and four board-certified non-nail expert dermatologists: 83.9% vs 73.0%; P = 0.005. The "NAPSI calculator" can be readily used in a clinical situation, contributing to raising the medical practice level for nail psoriasis.

Understanding the dynamic states and transitions of heterogeneous cell populations is crucial for addressing fundamental biological questions. High-content imaging provides rich datasets, but it remains increasingly difficult to integrate and annotate high-dimensional and time-resolved datasets to profile heterogeneous cell population dynamics in different microenvironments. Using hepatic stellate cells (HSCs) LX-2 as model, we proposed a novel analytical strategy for image-based integration and annotation to profile dynamics of heterogeneous cell populations in 2D/3D microenvironments. High-dimensional features were extracted from extensive image datasets, and cellular states were identified based on feature profiles. Time-series clustering revealed distinct temporal patterns of cell shape and actin cytoskeleton reorganization. We found LX-2 showed more complex membrane dynamics and contractile systems with an M-shaped actin compactness trend in 3D culture, while they displayed rapid spreading in early 2D culture. This image-based integration and annotation strategy enhances our understanding of HSCs heterogeneity and dynamics in complex extracellular microenvironments.

Amyotrophic Lateral Sclerosis (ALS) is a devastating, immensely complex neurodegenerative disease by lack of effective treatments. We developed a network medicine methodology via integrating human brain multi-omics data to prioritize drug targets and repurposable treatments for ALS. We leveraged non-coding ALS loci effects from genome-wide associated studies (GWAS) on human brain expression quantitative trait loci (QTL) (eQTL), protein QTL (pQTL), splicing QTL (sQTL), methylation QTL (meQTL), and histone acetylation QTL (haQTL). Using a network-based deep learning framework, we identified 105 putative ALS-associated genes enriched in known ALS pathobiological pathways. Applying network proximity analysis of predicted ALS-associated genes and drug-target networks under the human protein-protein interactome (PPI) model, we identified potential repurposable drugs (i.e., Diazoxide and Gefitinib) for ALS. Subsequent validation established preclinical evidence for top-prioritized drugs. In summary, we presented a network-based multi-omics framework to identify drug targets and repurposable treatments for ALS and other neurodegenerative disease if broadly applied.

Biomarkers associated with the progression from gastric intestinal metaplasia (GIM) to gastric adenocarcinoma (GA), i.e., GA-related GIM, could provide valuable insights into identifying patients with increased risk for GA. The aim of this study was to utilize multi-bioinformatics to reveal potential biomarkers for the GA-related GIM and predict potential drug repurposing for GA prevention in patients. The multi-bioinformatics included gene expression matrix (GEM) by microarray gene expression (MGE), ScType (a fully automated and ultra-fast cell-type identification based solely on a given scRNA-seq data), Ingenuity Pathway Analysis, PageRank centrality, GO and MSigDB enrichments, Cytoscape, Human Protein Atlas and molecular docking analysis in combination with immunohistochemistry. To identify GA-related GIM, paired surgical biopsies were collected from 16 GIM-GA patients who underwent gastrectomy, yielding 64 samples (4 biopsies per stomach x 16 patients) for MGE. Co-analysis was performed by including scRNAseq and immunohistochemistry datasets of endoscopic biopsies of 37 patients. The results of the present study showed potential biomarkers for GA-related GIM, including GEM of individual patients, individual genes (such as RBP2 and CD44), signaling pathways, network of molecules, and network of signaling pathways with key topological nodes. Accordingly, potential treatment targets with repurposed drugs were identified including epidermal growth factor receptor, proto-oncogene tyrosine-protein kinase Src, paxillin, transcription factor Jun, breast cancer type 1 susceptibility protein, cellular tumor antigen p53, mouse double minute 2, and CD44.

Rheumatoid Arthritis (RA) is a chronic autoimmune inflammatory disease that affects about 0.1% to 2% of the population worldwide. Despite the development of several novel therapies, there is only limited benefit for many patients. Thus, there is room for new approaches to improve response to therapy, including designing better trials e.g., by identifying subpopulations that can benefit from specific classes of therapy and enabling reverse translation by analyzing completed clinical trials. We have developed an open-source, mechanistic multi-scale model of RA, which captures the interactions of key immune cells and mediators in an inflamed joint. The model consists of a treatment-naive Virtual Population (Vpop) that responds appropriately (i.e. as reported in clinical trials) to standard-of-care treatment options-Methotrexate (MTX) and Adalimumab (ADA, anti-TNF-α) and an MTX inadequate responder sub-population that responds appropriately to Tocilizumab (TCZ, anti-IL-6R) therapy. The clinical read-outs of interest are the American College of Rheumatology score (ACR score) and Disease Activity Score (DAS28-CRP), which is modeled to be dependent on the physiological variables in the model. Further, we have validated the Vpop by predicting the therapy response of TCZ on ADA Non-responders. This paper aims to share our approach, equations, and code to enable community evaluation and greater adoption of mechanistic models in drug development for autoimmune diseases.

Chronic obstructive pulmonary disease (COPD) is an etiologically complex disease characterized by acute exacerbations and stable phases. We aimed to identify biological functions modulated in specific COPD conditions, using whole blood samples collected in the AERIS clinical study (NCT01360398). Considered conditions were exacerbation onset, severity of airway obstruction, and presence of respiratory pathogens in sputum samples. With an integrative multi-network gene community detection (MNGCD) approach, we analyzed expression profiles to identify communities of correlated genes. The approach combined different layers of gene interactions for each explored condition/subset of samples: gene expression similarity, protein-protein interactions, transcription factors, and microRNAs validated regulons. Heme metabolism, interferon-alpha, and interferon-gamma pathways were modulated in patients at both exacerbation and stable-state visits, but with the involvement of distinct sets of genes. An important gene community was enriched with G2M checkpoint, E2F targets, and mitotic spindle pathways during exacerbation. Targets of TAL1 regulator and hsa-let-7b - 5p microRNA were modulated with increasing severity of airway obstruction. Bacterial infections with Moraxella catarrhalis and, particularly, Haemophilus influenzae triggered a specific cellular and inflammatory response in acute exacerbations, indicating an active reaction of the host to infections. In conclusion, COPD is a complex multifactorial disease that requires in-depth investigations of its causes and features during its evolution and whole blood transcriptome profiling can contribute to capturing some relevant regulatory mechanisms associated with this disease. In this work, we explored multi-network modeling that integrated diverse layers of regulatory gene networks and enhanced our comprehension of the biological functions implicated in the COPD pathogenesis.

Elucidating the emergent dynamics of cellular differentiation networks is crucial to understanding cell-fate decisions. Toggle switch - a network of mutually repressive lineage-specific transcription factors A and B - enables two phenotypes from a common progenitor: (high A, low B) and (low A, high B). However, the dynamics of networks enabling differentiation of more than two phenotypes from a progenitor cell has not been well-studied. Here, we investigate the dynamics of a four-node network A, B, C, and D inhibiting each other, forming a toggle tetrahedron. Our simulations show that this network is multistable and predominantly allows for the co-existence of six hybrid phenotypes where two of the nodes are expressed relatively high as compared to the remaining two, for instance (high A, high B, low C, low D). Finally, we apply our results to understand naïve CD4⁺ T cell differentiation into Th1, Th2, Th17 and Treg subsets, suggesting Th1/Th2/Th17/Treg decision-making to be a two-step process.

Genome-scale metabolic models (GEMs) cover the entire list of metabolic genes in an organism and associated reactions, in a tissue/condition non-specific manner. RNA-seq provides crucial information to make the GEMs condition-specific. Integrative Metabolic Analysis Tool (iMAT) and Integrative Network Inference for Tissues (INIT) are the two most popular algorithms to create condition-specific GEMs from human transcriptome data. The normalization method of choice for raw RNA-seq count data affects the model content produced by these algorithms and their predictive accuracy. However, a benchmark of the RNA-seq normalization methods on the performance of iMAT and INIT algorithms is missing in the literature. Another important phenomenon is covariates such as age and gender in a dataset, and they can affect the predictivity of analysis. In this study, we aimed to compare five different RNA-seq data normalization methods (TPM, FPKM, TMM, GeTMM, and RLE) and covariate adjusted versions of the normalized data by mapping them on a human GEM using the iMAT and INIT algorithms to generate personalized metabolic models. We used RNA-seq data for Alzheimer's disease (AD) and lung adenocarcinoma (LUAD) patients. The results demonstrated that RNA-seq data normalized by the RLE, TMM, or GeTMM methods enabled the production of condition-specific metabolic models with considerably low variability in terms of the number of active reactions compared to the within-sample normalization methods (FPKM, TPM). Using these models, we could more accurately capture the disease-associated genes (average accuracy of ~0.80 for AD and ~0.67 for LUAD) for the RLE, TMM, and GeTMM normalization methods. An increase in the accuracies was observed for all the methods when covariate adjustment was applied. We found a similar accuracy trend when we compared the metabolites of perturbed reactions to metabolome data for AD. Together, our benchmark study shows that the between-sample RNA-seq normalization methods reduce false positive predictions at the expense of missing some true positive genes when mapped on GEMs.

Tumor hypoxia is a dynamic phenomenon marked by fluctuations in oxygen levels across both rapid (seconds to minutes) and slow (hours to days) time scales. While short hypoxia cycles are relatively well understood, the mechanisms behind longer cycles remain largely unclear. In this paper, we present a novel mechanistic mathematical model that explains slow hypoxia cycles through feedback loops involving vascular expansion and regression, oxygen-regulated tumor growth, and toxic cytokine production. Our study reveals that, for the emergence of slow hypoxia cycles, endothelial cells must adapt by decreasing receptor activation as ligand concentration increases. Additionally, the interaction between tumor cells and toxic cytokines influences frequency and intensity of these cycles. By examining the effects of pharmacological interventions, specifically poly (ADP-ribose) polymerase inhibitors, we also demonstrate how targeting cell proliferation can help regulate oxygen levels. Our findings enhance the understanding of hypoxia regulation and suggest PARP proteins as promising therapeutic targets.