Clinical Nutrition Research Centre

Accurate and efficient climate simulations are crucial for understanding Earth’s evolving climate. However, current General Circulation Models (GCMs) face challenges in capturing unresolved physical processes, such as cloud and convection. A common solution is to adopt Cloud-Resolving Models (CRMs), which provide more accurate results than the standard subgrid parameterization schemes typically used in GCMs. However, CRMs (also referred to as super-parameterizations, such as SPCAM) remain computationally prohibitive. Hybrid modeling, which integrates deep learning with equation-based GCMs, offers a promising alternative but often struggles with long-term stability and accuracy issues. In this work, we find that water vapor oversaturation during condensation is a key factor compromising the stability of hybrid modeling. To address this, we introduce CondensNet, a novel neural network architecture that embeds a self-adaptive physical constraint to correct unphysical condensation processes. CondensNet effectively mitigates water vapor oversaturation, enhancing simulation stability while maintaining accuracy and improving computational efficiency compared to super-parameterization schemes. We integrate CondensNet into a GCM to form PCNN-GCM (Physics-Constrained Neural Network GCM), a hybrid deep learning framework designed for long-term stable climate simulations under real-world conditions (AMIP setting). PCNN-GCM enables stable simulations over decades and achieves up to 370× speed-up compared with SPCAM, while also being faster than traditional CAM5 under GPU acceleration or CPU-only. Beyond stability and efficiency, PCNN-GCM demonstrates greater skill in capturing complex climate variability than CAM5, including tropical precipitation extremes and the Madden-Julian Oscillation (MJO), yielding results that align more closely with observations or reanalyses (e.g., ERA5, TRMM) than conventional parameterization schemes.

Great emphasis is placed upon addressing hyperinsulinism in at-risk infants with recurrent hypoglycemia. We report two preterm infants with recurrent nonhyperinsulinemic hypoglycemia due to transient impairment of glucose mobilization, warranting diazoxide use. Inadequate response to glucagon stimulation test indicates hepatic enzyme immaturity especially with negative genetic testing for glycogen storage disorders.

s 1 2 9 5 1 -0 2 5 -0 3 2 0 2 -7In this article, the cell uptake time points 1 h, 3 h, 6 h, and 9 h were incorrectly labeled as control, 1 h, 3 h, and 6 h in Fig. 3a.Fig 3 appeared incorrectly and have now been corrected in the original publication.For completeness and transparency, the correct and incorrect versions are displayed below.

Inflammation significantly affects the progression of gastric cancer (GC). The Systemic Inflammation Response Index (SIRI) is an emerging prognostic marker for various malignant tumors, including GC. However, the mechanisms by which SIRI influences patient outcomes remain unclear. This study employed a multi-omics approach to analyze GC patients with different SIRI levels, aiming to identify factors underlying differences in prognosis. A cohort of 495 GC patients was classified into high SIRI (n = 276) and low SIRI (n = 219) groups using a threshold of 2.6. Kaplan-Meier analysis and Receiver operating characteristic (ROC) curves were used to assess survival outcomes and the predictive accuracy of SIRI compared to other inflammatory markers. Gut microbiota and untargeted metabolomics analyses were conducted on stool samples from 45 GC patients, including 21 with high SIRI and 24 with low SIRI. Elevated SIRI levels were significantly associated with worse OS (P = 0.004, HR = 1.845, 95% CI: 1.231-2.766) and DFS (P < 0.0001, HR = 3.102, 95% CI: 2.016-4.772) in GC patients. Multi-omics analysis revealed distinct gut microbial and metabolic profiles between high- and low-SIRI subgroups. High SIRI was linked to increased Escherichia-Shigella and decreased levels of Blautia and Klebsiella. Metabolomic differences included elevated N-Acetylcadaverine and Epsilon-caprolactam, and decreased S-(PGA2)-glutathione. Integrative analysis identified significant microbe-metabolite correlations, such as Escherichia-Shigella with N-Acetylcadaverine and Epsilon-caprolactam. These findings suggest that SIRI-associated microbial and metabolic features may serve as non-invasive markers for inflammatory risk stratification in GC. High SIRI levels are associated with poorer prognosis in GC patients. Significant differences in gut microbiota and metabolites were observed across different SIRI levels in GC patients, showing their potential as non-invasive markers for GC risk stratification based on SIRI levels.

Abstract Collecting duct carcinoma (CDC) is an aggressive rare subtype of kidney cancer with unmet clinical needs. Little is known about its underlying molecular alterations and etiology, primarily due to its rarity, and lack of preclinical models. This study aims to comprehensively characterize molecular alterations in CDC and identify its therapeutic vulnerabilities. Through whole-exome and transcriptome sequencing, we identified KRAS hotspot mutations (G12A/D/V) in 3/13 (23%) of the patients, in addition to known TP53, NF2 mutations. 3/13 (23%) patients carried a mutational signature (SBS22) caused by aristolochic acid (AA) exposures, known to be more prevalent in Asia, highlighting a geologically specific disease etiology. We further discovered that cell cycle-related pathways were the most predominantly dysregulated pathways. Our drug screening with our newly established CDC preclinical models identified a CDK9 inhibitor LDC000067 that specifically inhibited CDC tumor growth and prolonged survival. Our study not only improved our understanding of oncogenic molecular alterations of Asian CDC, but also identified cell-cycle machinery as a therapeutic vulnerability, laying the foundation for clinical trials to treat patients with such aggressive cancer. Citation Format: Peiyong Guan, Jianfeng Chen, Chengqiang Mo, Fukawa Tomoya, Chao Zhang, Xiuyu Cai, Mei Li, Jing Han Hong, Jason Yongsheng Chan, Cedric Chuan Young Ng, Jing Yi Lee, Suet Far Wong, Wei Liu, Xian Zeng, Peili Wang, Rong Xiao, Vikneswari Rajasegaran, Swe Swe Myint, Abner Ming Sun Lim, Joe Poh Sheng Yeong, Puay Hoon Tan, Choon Kiat Ong, Tao Xu, Yiqing Du, Fan Bai, Xin Yao, Bin Tean Teh, Jing Tan. Comprehensive Molecular Characterization of Collecting Duct Carcinoma for Therapeutic Vulnerability [abstract]. In: Proceedings of Frontiers in Cancer Science 2024; 2024 Nov 13-15; Singapore. Philadelphia (PA): AACR; Cancer Res 2025;85(15_Suppl):Abstract nr P03.

Insulin resistance (IR) is fundamental to the development of type 2 diabetes and usually precedes the disease by several years. Therefore, a better understanding of the molecular pathogenesis of IR may help better manage and prevent the metabolic syndrome. We used an integrated approach of anthropometric and biochemical measures in conjunction with serum lipidomics for metabolic phenotyping of multiethnic male participants with varied insulin sensitivity. By detailed characterization of the serum lipidomic profiles in the fasting state as well as temporal changes during the hyperinsulinemic-euglycemic clamp procedure (HIEC), we identified a systemic metabolic response to the HIEC represented by the marked changes in the acylcarnitine, nonesterified fatty acid, lysophospholipid, sphingosine-1-phosphate, and phosphatidylserine lipid classes. We demonstrate that a shared lipidomic signature between IR and liver fat shows gradual increase with increasing liver fat percentage. In summary, by stratifying the study participants according to the insulin sensitivity indices derived from the HIEC procedures, we identified a circulating lipidomic signature of IR and metabolic plasticity with a potential for prediction and management of metabolic health before the development of type 2 diabetes and other metabolic diseases.Article HighlightsThe underlying molecular pathogenesis of the Asian phenotype of insulin resistance remains to be understood.We carried out metabolic phenotyping of study participants without diabetes according to insulin sensitivity indices derived from hyperinsulinemic-euglycemic clamp procedures.We identified lipidomic signatures of insulin resistance and metabolic plasticity.These lipidomic signatures have the potential to help in risk stratification of insulin resistance and metabolic dysfunction for early intervention.

Human beings may have evolved the largest asymmetries of brain organization in the animal kingdom. Hemispheric left-vs-right specialization is especially pronounced in species-unique capacities, including emotional processing such as facial judgments, language-based feats such as reading books, and creativity such as musical performances. We hence chart the largest longitudinal brain-imaging resource, and provide evidence that brain asymmetry changes continuously in a manner suggestive of neural plasticity throughout adulthood. In the UK Biobank population cohort, we demonstrate that whole-brain patterns of asymmetry changes show robust phenome-wide associations across 959 distinct variables spanning 11 categories. We also find that changes in brain asymmetry over years co-occur with changes among specific lifestyle markers. We uncover specific brain asymmetry changes which systematically co-occur with entering a new phase of life, namely retirement. Finally, we reveal relevance of evolving brain asymmetry within subjects to major disease categories across ~4500 total medical diagnoses. Our findings speak against the idea that asymmetrical neural systems are conserved throughout adulthood.

Abstract. Land cover and land management changes (LCLMCs) have a substantial impact on the global carbon budget and, consequently, on global climate via the biogeochemical (BGC) effects. The commonly considered BGC effects refer to the direct influence of LCLMCs on local carbon stocks (local BGC effects). However, LCLMCs also influence climate by altering the local surface energy balance due to changes in land surface properties, such as albedo, leaf area, and roughness (local biogeophysical (BGP) effects). Altered local air mass properties can impact regions remote from LCLMCs through advection and changes in large-scale circulation (nonlocal BGP effects). Previous studies have shown potentially substantial nonlocal BGP effects on temperature and precipitation. Given that the terrestrial carbon cycle strongly depends on climate conditions, this raises the question of whether LCLMCs can trigger remote carbon cycle changes (nonlocal BGC effects) – a currently overlooked, potentially large climate and ecosystem impact. To assess the nonlocal BGC effects, we analyze sensitivity simulations for three selected types of hypothetical large-scale LCLMCs, global cropland expansion, global cropland expansion with irrigation, and global afforestation, which were performed by three state-of-the-art Earth system models (ESMs). We separate the nonlocal BGC effect using a checkerboard-like LCLMC perturbation that has previously only been applied to BGP effects. We show that nonlocal BGC effects on vegetation and soil carbon pools persistently accumulate, exceeding natural fluctuations and typically becoming detectable within the first 40 years after LCLMCs. By the end of our 160-year simulation period, nonlocal BGC effects lead to an absolute magnitude of change in total terrestrial carbon stock by 1 to 37 GtC, with strong changes over the densely forested Amazon region (0.2 to 7 GtC) and central Congo Basin region (0.3 to 15 GtC), depending on models and LCLMCs implemented. For the irrigation scenario, the nonlocal BGC effects are comparable to the total BGC effects, with the nonlocal-to-total ratio for vegetation carbon pools commonly reaching around 90 %. Our results reveal that the nonlocal BGC effects could be substantial and call for these effects to be considered for accurate impact assessment and sound policymaking. This becomes even more relevant when LCLMCs are expected to play a pivotal role in achieving the Paris Agreement's goal of limiting global warming below 1.5 °C above pre-industrial levels.

A pervasive dilemma in brain-wide association studies1 (BWAS) is whether to prioritize functional magnetic resonance imaging (fMRI) scan time or sample size. We derive a theoretical model showing that individual-level phenotypic prediction accuracy increases with sample size and total scan duration (sample size × scan time per participant). The model explains empirical prediction accuracies well across 76 phenotypes from nine resting-fMRI and task-fMRI datasets (R2 = 0.89), spanning diverse scanners, acquisitions, racial groups, disorders and ages. For scans of ≤20 min, accuracy increases linearly with the logarithm of the total scan duration, suggesting that sample size and scan time are initially interchangeable. However, sample size is ultimately more important. Nevertheless, when accounting for the overhead costs of each participant (such as recruitment), longer scans can be substantially cheaper than larger sample size for improving prediction performance. To achieve high prediction performance, 10 min scans are cost inefficient. In most scenarios, the optimal scan time is at least 20 min. On average, 30 min scans are the most cost-effective, yielding 22% savings over 10 min scans. Overshooting the optimal scan time is cheaper than undershooting it, so we recommend a scan time of at least 30 min. Compared with resting-state whole-brain BWAS, the most cost-effective scan time is shorter for task-fMRI and longer for subcortical-to-whole-brain BWAS. In contrast to standard power calculations, our results suggest that jointly optimizing sample size and scan time can boost prediction accuracy while cutting costs. Our empirical reference is available online for future study design (https://thomasyeolab.github.io/OptimalScanTimeCalculator/index.html).