Genomic and Epigenomic Profiling of Cancer

Cancer is a genetic disease that arises from cells undergoing genomic alterations. Understanding the role of genomic instability in tumorigenesis, progression, and metastasis is crucial for advancing cancer diagnosis, treatment, and drug development. The development of cancer is driven by a combination of genetic variations, epigenetic dysregulation, and environmental influences. Among the most aggressive malignancies worldwide, gastroesophageal cancer and glioblastoma are characterized by extremely poor prognoses and limited therapeutic options. We explored GCA from a genomic perspective, integrating whole exome sequencing, RNAseq, proteomics, and metabolomics to identify key genetic alterations and signaling pathways that drive tumorigenesis. Additionally, we investigated the epigenetic landscape of glioblastoma to reveal the role of epigenetic dysregulation in tumor heterogeneity and progression. In Paper I, we performed whole genome sequencing on 36 pairs of tumor and tumor-matched normal samples from a GCA cohort and conducted immunohistochemistry of HER2 in 1668 GCA patients. We found that focal amplifications were detected in 77.8% of all cases, while ecDNAs were identified in 52.8% of total cases. Surprisingly, we found patients with ERBB2 focal amplification or IHC HER2 positive staining are associated with better prognosis, which is inconsistent with many of the previous studies that the oncogene ERBB2 typically correlates with poorer prognosis in patients. In Paper II, we conducted multi-omics profiling of 128 GCA patients, categorizing them into HER2-high, HER2-low, and HER2-negative groups. HER2 was identified as a favorable prognostic marker, with DNA repair features enriched in the HER2-high group and inflammation predominant in the HER2-low and HER2-negative groups. ARID1A mutations were particularly prognostic in the HER2-negative group. Our findings suggest antiinflammatory therapies and CD47/SIRPA immune checkpoint inhibition as potential strategies for HER2-negative GCA, offering new avenues for personalized treatment. In Paper III, we integrated analyses of chromatin accessibility, histone modifications (H3K4me1, H3K4me3, H3K27ac, H3K27me3), and chromatin loops in human and mouse GSCs. We found the enhancer marker H3K4me1 and the repressive marker H3K27me3 in human GSCs could separate patients into two groups with significant survival differences and enhancer signatures that define glioblastoma stem cell subtypes. Transcription factor