Highlights d Propagation of variation (PoV) in gene expression depends on biochemical parameters d Different cytokine environments can tune system parameters to affect PoV d Transcription factor activity and interaction with target enhancers can tune PoV d Natural perturbation of cellular and biochemical parameters can reveal PoV regulators
BACKGROUND: Aspiration pneumonia after endoscopic submucosal dissection (ESD) is rare, but can be fatal. We aimed to investigate risk factors and develop a simple risk scoring system for aspiration pneumonia. METHODS: We retrospectively reviewed medical records of 7833 patients who underwent gastric ESD for gastric neoplasm under anesthesiologist-directed sedation. Candidate risk factors were screened and assessed for significance using a least absolute shrinkage and selection operator (LASSO)-based method. Top significant factors were incorporated into a multivariable logistic regression model, whose prediction performance was compared with those of other machine learning models. The final risk scoring system was created based on the estimated odds ratios of the logistic regression model. RESULTS: The incidence of aspiration pneumonia was 1.5%. The logistic regression model showed comparable performance to the best predictive model, extreme gradient boost (area under receiver operating characteristic curve [AUROC], 0.731 vs 0.740). The estimated odds ratios were subsequently used for the development of the clinical scoring system. The final scoring system exhibited an AUROC of 0.730 in the test dataset with risk factors: age (≥70 years, 4 points), male sex (8 points), body mass index (≥27 kg/m 2 , 4 points), procedure time (≥80 minutes, 5 points), lesion in the lower third of the stomach (5 points), tumor size (≥10 mm, 3 points), recovery time (≥35 minutes, 4 points), and desaturation during ESD (9 points). For patients with total scores ranging between 0 and 33 points, aspiration pneumonia probabilities spanned between 0.1% and 17.9%. External validation using an additional cohort of 827 patients yielded AUROCs of 0.698 for the logistic regression model and 0.680 for the scoring system. CONCLUSIONS: Our simple risk scoring system has 8 predictors incorporating patient-, procedure-, and sedation-related factors. This system may help clinicians to stratify patients at risk of aspiration pneumonia after ESD. (Anesth Analg 2022;134:114-22) KEY POINTS• Question: What are the possible predictors for a risk scoring system of aspiration pneumonia after endoscopic submucosal dissection (ESD)? • Finding: Our risk scoring system has 8 predictors; older age, male sex, higher body mass index (BMI), longer procedure time, lesion located in the lower third of the stomach, larger tumor size, prolonged recovery time, and desaturation during ESD. • Meaning: To our knowledge, this study was the first to develop a risk scoring system for aspiration pneumonia after gastric ESD which enables clinicians to stratify patients into different risk categories. GLOSSARYASA = American Society of Anesthesiologists; AUROC = area under receiver operating characteristic curve; BMI = body mass index; CI = confidence interval; COPD = chronic obstructive pulmonary disease; CV = cross-validation; ESD = endoscopic submucosal dissection; IPA = index of prediction accuracy; L3 = the third lumbar vertebra; LASSO = least absolute shrinkage and select...
The majority of genetic variants associated with complex human autoimmune diseases reside in enhancers1–3, non-coding regulatory elements that control gene expression. In contrast with variants that directly alter protein-coding sequences, enhancer variants are predicted to tune gene expression modestly and function in specific cellular contexts4, suggesting that small alterations in the functions of key immune cell populations are sufficient to shape disease risk. Here we tested this concept by experimentally perturbing distinct enhancers governing the high affinity IL-2 receptor alpha chain (IL2RA; also known as CD25). IL2RA is an immune regulator that promotes the pro- and anti-inflammatory functions of conventional T cells (Tconvs) and regulatory T cells (Tregs), respectively, and non-coding genetic variants in IL2RA have been linked to multiple autoimmune disorders4. We previously tiled across the IL2RA locus using CRISPR-activation and identified a stimulation-responsive element (CaRE4) with an enhancer that modestly affects the kinetics of IL2RA expression in Tconvs5. This enhancer is conserved across species and harbors a common human SNP associated with protection from Type 1 Diabetes (T1D)5,6. We now identified an additional conserved enhancer, termed CaRE3 enhancer, which modestly affected steady state IL2RA expression in regulatory T cells (Tregs). Despite their seemingly subtle impact on gene expression, the CaRE3 and CaRE4 enhancers had pronounced yet divergent effects on the incidence of diabetes in autoimmune prone animals. Deletion of the conserved CaRE4 enhancer completely protected against autoimmune diabetes even in animals treated with an immunostimulating anti-PD1 checkpoint inhibitor, whereas deletion of the CaRE3 enhancer accelerated spontaneous disease progression. Quantitative multiplexed imaging of the pancreatic lymph nodes (panLNs) revealed that each enhancer deletion preferentially affected the protein expression levels of IL2RA in activated Tconvs or Tregs, reciprocally tuning local competition for IL-2 input signals. In animals lacking the CaRE4 enhancer, skewed IL-2 signaling favored Tregs, increasing their local density around activated Tconvs to strongly suppress emergence of autoimmune effectors. By contrast, in animals lacking the CaRE3 enhancer, IL-2 signals were skewed towards activated Tconvs, promoting their escape from Treg control. Collectively, this work illustrates how subtle changes in gene regulation due to non-coding variation can significantly alter disease progression and how distinct enhancers controlling the same gene can have opposing effects on disease outcomes through cell type-selective activity.
15A recurrent challenge in biology is the development of predictive quantitative models 16 because most molecular and cellular parameters have unknown values and realistic models 17 are analytically intractable. While the dynamics of the system can be analyzed via 18 computer simulations, substantial computational resources are often required given 19uncertain parameter values resulting in large numbers of parameter combinations, 20 especially when realistic biological features are included. Simulation alone also often does 21 not yield the kinds of intuitive insights from analytical solutions. Here we introduce a 22A major goal of systems biology is to develop quantitative models to predict the behavior 36 of biological systems (1, 2). However, most realistic molecular and cellular models have a 37 large number of parameters (e.g., reaction rates, cellular proliferation rates, extent of 38 physical interactions among cells or molecules), whose values remain unknown and are 39 often challenging to measure or infer quantitatively (3, 4). While some biological 40 phenotypes are robust to parameter variations (5), most are "tunable" by parameters (6). 41Therefore, analyzing the behavior of a system over the entire plausible space of parameters 42 is needed to study the phenotypic range of a biological system and its parameter-phenotype 43 relationships (7-9). A case in point concerns a contemporary problem in single cell biology: 44Despite the increasing availability of single-cell gene expression data enabled by rapid 45 technological advances (10), an important unanswered question is how cell-to-cell 46 expression variation and gene-gene correlation among single cells are regulated by the 47 computational resources are required to analyze the parameter space, given the large 59 uncertainty in parameter values and the complex correlation structure among parameters. 60Plus, simulation analysis alone often does not automatically yield intuitive understanding. 61Here, we combine computational simulation of full-feature stochastic models and 62 machine learning (ML) to develop a framework, called MAchine learning of Parameter-63Phenotype Analysis (MAPPA), for constructing, exploring, and analyzing the mapping 64 between parameters and quantitative phenotypes of a stochastic dynamical system (Figures 65 1A and S1; Supplementary text). Our goal is to take advantage of the large amounts of data 66that can be generated from bottom-up, mechanistic computational simulation of dynamical 67 systems and the ability of modern machine learning approaches to "compress" such data 68 to generate computationally efficient and interpretable models. MAPPA thus builds 69 efficient, predictive, and interpretable ML models that capture the nonlinear mapping 70 between parameter and phenotypic spaces (parameter-phenotype maps). The ML models 71 can be viewed as "phenomenological" solutions of the SME that can predict the system's 72 quantitative behavior from parameter combinations, thus bypassing computationally 73 expensive simulatio...
A diabetic patient may suffer simultaneously from cardiovascular disease; thus, lipid-lowering or anti-hypertensive agents could be given together with nateglinide. The pharmacokinetics of nateglinide were investigated in the presence and absence of HMG-CoA reductase inhibitors (fluvastatin, lovastatin) and calcium channel blockers (verapamil, nifedipine) in rabbits. A pharmacokinetic modeling approach was used to quantify the effects of the drugs that significantly influenced the pharmacokinetics of nateglinide. Fluvastatin and nifedipine shifted the time course of serum nateglinide concentrations upwards; there was no significant change with verapamil or lovastatin. The C(max) and AUC(inf) increased 1.5- (p<0.05) and 1.3-fold in the presence of fluvastatin and 1.8- (p<0.01) and 2.4-fold (p<0.01) in the presence of nifedipine, respectively. In a simultaneous nonlinear regression, fluvastatin and nifedipine decreased the elimination rate constant, by 76% and 32%, respectively. Fluvastatin and nifedipine increased the systemic exposure of nateglinide in rabbits, probably due to their inhibitory action on the metabolism of nateglinide by CYP2C5 (human CYP2C9). The concomitant use of fluvastatin and/or nifedipine with nateglinide is quite likely; therefore, the clinical consequences of long-term treatments must be considered.
Background Several predictive factors for chronic kidney disease (CKD) following radical nephrectomy (RN) or partial nephrectomy (PN) have been identified. However, early postoperative laboratory values were infrequently considered as potential predictors. Therefore, this study aimed to develop predictive models for CKD 1 year after RN or PN using early postoperative laboratory values, including serum creatinine (SCr) levels, in addition to preoperative and intraoperative factors. Moreover, the optimal SCr sampling time point for the best prediction of CKD was determined. Methods Data were retrospectively collected from patients with renal cell cancer who underwent laparoscopic or robotic RN (n = 557) or PN (n = 999). Preoperative, intraoperative, and postoperative factors, including laboratory values, were incorporated during model development. We developed 8 final models using information collected at different time points (preoperative, postoperative day [POD] 0 to 5, and postoperative 1 month). Lastly, we combined all possible subsets of the developed models to generate 120 meta-models. Furthermore, we built a web application to facilitate the implementation of the model. Results The magnitude of postoperative elevation of SCr and history of CKD were the most important predictors for CKD at 1 year, followed by RN (compared to PN) and older age. Among the final models, the model using features of POD 4 showed the best performance for correctly predicting the stages of CKD at 1 year compared to other models (accuracy: 79% of POD 4 model versus 75% of POD 0 model, 76% of POD 1 model, 77% of POD 2 model, 78% of POD 3 model, 76% of POD 5 model, and 73% in postoperative 1 month model). Therefore, POD 4 may be the optimal sampling time point for postoperative SCr. A web application is hosted at https://dongy.shinyapps.io/aki_ckd. Conclusions Our predictive model, which incorporated postoperative laboratory values, especially SCr levels, in addition to preoperative and intraoperative factors, effectively predicted the occurrence of CKD 1 year after RN or PN and may be helpful for comprehensive management planning.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.