Somatic Mutation-Informed Transcriptional Architecture of Hepatocellular Carcinoma: A Multiomics Study Centered on TP53.

Hepatocellular carcinoma (HCC) is characterized by extensive somatic mutation heterogeneity, yet the cellular contexts in which recurrent driver mutation-associated transcriptional programs operate remain incompletely defined. In this study, we applied a mutation-informed multiomics framework centered on TP53 to connect somatic mutation status, bulk transcriptional programs, and hepatocyte-like cell states at single-cell resolution. Somatic mutation profiles and transcriptomic data from the TCGA-LIHC cohort were used to construct directional mutation-associated transcriptional signatures, with CTNNB1 retained as a reference driver signature and AXIN1 and ARID1A retained for descriptive comparison. These bulk-derived transcriptional programs were then projected into single-cell RNA sequencing data to evaluate inferred TP53-associated transcriptional activity across hepatocyte-like populations. TP53 mutation-associated transcriptional programs were enriched in proliferation-related pathways, DNA repair processes, and stress-response programs and were associated with adverse clinical outcomes. At the single-cell level, inferred TP53-associated transcriptional activity showed heterogeneous distribution across hepatocyte-like populations and was preferentially enriched in putative malignant hepatocyte-like states. Pseudobulk and functional program analyses further supported the association between TP53-high signature states, cell-cycle activity, DNA repair, stress programs, and altered tumor-cell communication patterns. Integrated prioritization identified CDCA8, MCM10, and TRIP13 as representative genes associated with TP53-related transcriptional programs; these genes should be interpreted as components of a TP53-associated proliferative transcriptional state rather than as experimentally validated drivers. These findings provide a mutation-informed view of transcriptional architecture in HCC and illustrate how bulk mutation-associated signatures can be integrated with single-cell data to improve biological interpretation of somatic variation.