AI and Cancer Research ── Cancer biology ── 2026-10-31
Cancer molecular biology and AI
- Abstract
Background: Metabolic dysfunction-associated steatohepatitis (MASH) increases the risk of cirrhosis and liver cancer, with liver fibrosis being the primary determinant of adverse outcomes. However, the molecular mechanisms driving fibrosis progression in MASH remains incompletely understood.
Methods: We integrated four bulk RNA-seq datasets into a merged cohort (MergeCohort) and analyzed one single-nucleus RNA-seq (snRNA-seq) dataset. Robust Rank Aggregation (RRA), WGCNA and machine learning were combined to identify core genes associated with MASH fibrosis. Additionally, we established MASH mice model induced by methionine-choline-deficient (MCD) diet, and further check the expression of hub genes through quantitative real-time PCR (qRT-PCR). Single-nucleus transcriptomic profiling of liver tissues from MASH patients across distinct fibrosis stages was investigated to delineate cellular heterogeneity, reconstruct intercellular communication networks, and characterize key transcription factors (TFs) as well as receptor-ligand interactions involved in MASH-associated fibrosis.
Results: Based on the MergeCohort, we detected two key gene network modules associated with fibrosis progression in MASH through WGCNA. By integrating machine learning algorithms, we discovered seven core genes (CAPN6, FAP, KRT23, SOX4, STMN2, TAGLN and VWF) showing favorable discriminatory capability for fibrosis stage, achieving an area under the curve (AUC) of 0.81 (95% CI 0.69–0.93) in a distinct external cohort. Moreover, CAPN6, FAP and KRT23 showed directionally consistent upregulation in the liver tissues of MASH mouse model induced by MCD diet. Subsequently, applying cell-type deconvolution analysis, we assigned the predominant cell types (hepatic stellate cells (HSCs), cholangiocytes, hepatocytes and macrophages) to the key WGCNA modules. At the single-nucleus level, we identified FGF13 + cholangiocytes and SAAs + hepatocytes subpopulations. Transcription factor regulon analysis identified key TFs (CREB5 and STAT3), which may play potential regulatory roles in FGF13 + cholangiocytes and SAAs + hepatocytes. Cell-cell communication analysis further revealed crucial ligand-receptor pairs between HSCs and FGF13 + cholangiocytes (VEGFA_NRP1 and APP_CD74 axes), SAAs + hepatocytes and HSCs (RSPO3_LGR4 and IGF1_IGF1R axes), SAAs + hepatocytes and macrophages (TNFSF14_TNFRSF14 and SELE_CD44 axes) associated with the progression of liver fibrosis in MASH.
Conclusion: This combined approach defines molecular signatures related to MASH fibrosis that may inform the development of novel therapeutic strategies and serve as a tissue signature associated with advanced fibrosis in MASH.
Journal IF-equivalent: 4.8 (OpenAlex 2-year mean citedness, value as of 2026-10-15, retrieved 2026-10-16; not the official Clarivate IF)Reference: Zhang J, Wang J, Shen Y, Zhang Y, Luo Z, Kong X, et al. Integrative machine learning analysis of bulk and single-nucleus RNA sequencing reveals key cell types and biomarkers driving MASH-associated fibrosis. Front Cell Dev Biol. 2026 Oct 29;14:1961567. doi:10.3389/fcell.2026.1961567.Checked: Abstract only - Abstract
BACKGROUND: Recent investigations have demonstrated the relevance in cancer of a variety of small RNAs, beyond microRNAs (miRNAs). However, the comprehensive application of these RNAs to cancer detection in dogs remains unexplored. HYPOTHESIS/OBJECTIVES: We investigated whether combining miRNAs and other small RNAs using a custom database improves the accuracy of machine learning cancer screening in dogs compared with models using only miRNAs. ANIMALS: A total of 389 dogs were enrolled: 85 with cancer, 77 with nonmalignant diseases, and 227 healthy controls. METHODS: Small RNAs in blood samples were sequenced, and miRNAs (via miRBase) and other small RNAs (via a custom database) were profiled. Based on these profiles, 2 datasets were prepared: one with only miRNAs and another combining miRNAs with other small RNAs. For each dataset, 12 classification models were constructed using identical methods to enable paired comparisons of performance, with 5-fold cross-validation and area under the receiver operating characteristic curve (AUC) as the primary metrics. RESULTS: The model that integrated both miRNAs and other small RNAs yielded a significantly higher AUC compared with the miRNA-only model (median AUC: 0.926 [95% CI, 0.911-0.944] vs 0.881 [0.860-0.897]; Wilcoxon signed-rank test, P < .001), indicating enhanced capability in discriminating malignant tumors in dogs. CONCLUSIONS AND CLINICAL IMPORTANCE: We investigated the comprehensive integration of diverse small RNA species, beyond miRNAs, using a machine learning approach for cancer screening in dogs. Although further prospective validation in independent cohorts is needed, our results indicate that this strategy improves the performance of cancer classification models in dogs.
Journal IF-equivalent: 3.4 (OpenAlex 2-year mean citedness, value as of 2026-10-29, retrieved 2026-10-31; not the official Clarivate IF)Reference: Harako S, Yuki M, Noguchi S, Ohno T, Nozaki N, Nishida R, et al. Novel liquid biopsy based on comprehensive small RNA profiling for screening of cancer in dogs utilizing machine learning. Journal of Veterinary Internal Medicine. 2026 Oct 30;40(5):aalag270. doi:10.1093/jvimsj/aalag270.Checked: Abstract only