A primary challenge in single-cell RNA sequencing (scRNA-seq) studies comes from the massive amount of data and the excess noise level. To address this challenge, we introduce an analysis framework, named single-cell Decomposition using Hierarchical Autoencoder (scDHA), that reliably extracts representative information of each cell. The scDHA pipeline consists of two core modules. The first module is a non-negative kernel autoencoder able to remove genes or components that have insignificant contributions to the part-based representation of the data. The second module is a stacked Bayesian autoencoder that projects the data onto a low-dimensional space (compressed). To diminish the tendency to overfit of neural networks, we repeatedly perturb the compressed space to learn a more generalized representation of the data. In an extensive analysis, we demonstrate that scDHA outperforms state-of-the-art techniques in many research sub-fields of scRNA-seq analysis, including cell segregation through unsupervised learning, visualization of transcriptome landscape, cell classification, and pseudo-time inference.

Hepatocellular carcinoma (HCC) is a commonly diagnosed malignancy with poor prognosis. Rising incidence of HCC may be due to rising prevalence of metabolic dysfunction-associated fatty liver disease, where altered bile acid metabolism may be implicated in HCC development. Thirty-five bile acids were quantified using ultra-performance liquid chromatography triple-quadrupole mass spectrometry assays in pre-diagnostic serum of 100 HCC cases and 100 matched controls from the Singapore Chinese Health Study. Conditional logistic regression was used to assess associations for bile acid levels with risk of HCC. Conjugated primary bile acids were significantly elevated whereas the ratios of secondary bile acids over primary bile acids were significantly lower in HCC cases than controls. The respective odds ratios and 95% confidence intervals of HCC were 6.09 (1.75–21.21) for highest vs. lowest tertile of cholic acid species and 30.11 (5.88–154.31) for chenodeoxycholic acid species. Doubling ratio of taurine-over glycine-conjugated chenodeoxycholic acid was associated significantly with 40% increased risk of HCC whereas doubling ratio of secondary over primary bile acid species was associated with 30–40% reduced risk of HCC. In conclusion, elevated primary bile acids and taurine over glycine-conjugated ratios were strongly associated with HCC risk whereas the ratios of secondary bile acids over primary bile acids were inversely associated with HCC risk.

With over 1 million incidence cases and more than 780,000 deaths in 2018, gastric cancer (GC) was ranked as the 5th most common cancer and the 3rd leading cause of cancer deaths worldwide. Though several biomarkers, including carcinoembryonic antigen (CEA), cancer antigen 19-9 (CA19-9), and cancer antigen 72-4 (CA72-4), have been identified, their diagnostic accuracies were modest. Circulating tumor cells (CTCs), cells derived from tumors and present in body fluids, have recently emerged as promising biomarkers, diagnostically and prognostically, of cancers, including GC. In this review, we present the landscape of CTCs from migration, to the presence in circulation, biologic properties, and morphologic heterogeneities. We evaluated clinical implications of CTCs in GC patients, including diagnosis, prognosis, and therapeutic management, as well as their application in immunotherapy. On the one hand, major challenges in using CTCs in GC were analyzed, from the differences of cut-off values of CTC positivity, to techniques used for sampling, storage conditions, and CTC molecular markers, as well as the unavailability of relevant enrichment and detection techniques. On the other hand, we discussed future perspectives of using CTCs in GC management and research, including the use of circulating tumor microembolies; of CTC checkpoint blockade in immunotherapy; and of organoid models. Despite the fact that there are remaining challenges in techniques, CTCs have potential as novel biomarkers and/or a non-invasive method for diagnostics, prognostics, and treatment monitoring of GC, particularly in the era of precision medicine.

Importance: The global burden of obesity is increasing, as are colorectal cancer (CRC) incidence and mortality.Objectives: To assess the association between body mass index (BMI) and risks of incident CRC and CRC-related death in the Asian population.Design, Setting, and Participants: This cohort study includes data pooled from 17 prospective cohort studies included in The Asia Cohort Consortium. Cohort enrollment was conducted from January 1, 1984, to December 31, 2002. Median follow-up time was 15.2 years (IQR, 12.1-19.2 years). Data were analyzed from January 15, 2023, through January 15, 2024.Exposure: Body mass index, calculated as weight in kilograms divided by height in meters squared.Main Outcomes and Measures: The primary outcomes were CRC incidence and CRC-related mortality. The risk of events is reported as adjusted hazard ratios (AHRs) and 95% CIs for incident CRC and death from CRC using the Cox proportional hazards regression model.Results: To assess the risk of incident CRC, 619 981 participants (mean [SD] age, 53.8 [10.1] years; 52.0% female; 11 900 diagnosed incident CRC cases) were included in the study, and to assess CRC-related mortality, 650 195 participants (mean [SD] age, 53.5 [10.2] years; 51.9% female; 4550 identified CRC deaths) were included in the study. A positive association between BMI and risk of CRC was observed among participants with a BMI greater than 25.0 to 27.5 (AHR, 1.09 [95% CI, 1.03-1.16]), greater than 27.5 to 30.0 (AHR, 1.19 [95% CI, 1.11-1.29]), and greater than 30.0 (AHR, 1.32 [95% CI, 1.19-1.46]) compared with those with a BMI greater than 23.0 to 25.0 (P < .001 for trend), and BMI was associated with a greater increase in risk for colon cancer than for rectal cancer. A similar association between BMI and CRC-related death risk was observed among participants with a BMI greater than 27.5 (BMI >27.5-30.0: AHR, 1.18 [95% CI, 1.04-1.34]; BMI >30.0: AHR, 1.38 [95% CI, 1.18-1.62]; P < .001 for trend) and was present among men with a BMI greater than 30.0 (AHR, 1.87 [95% CI, 1.49-2.34]; P < .001 for trend) but not among women (P = .15 for trend) (P = .02 for heterogeneity).Conclusions and Relevance: In this cohort study that included a pooled analysis of 17 cohort studies comprising participants across Asia, a positive association between BMI and CRC incidence and related mortality was found. The risk was greater among men and participants with colon cancer. These findings may have implications to better understand the burden of obesity on CRC incidence and related deaths in the Asian population.