Trung Nghia Vu

The overall aim is to utilize multi-omics approach to identify novel etiopathogenesis and early detection biomarkers for stomach cancer and its precursor lesions. To achieve this aim, first we will use stored serum samples to perform metabolomics profiling among 12,599 twin subjects, among whom 1034 were deemed to have chronic atrophic gastritis based on measured pepsinogen I and II levels. Logistic regression will be used to search for metabolites related to the risk of chronic atrophic gastritis. Second, we will further measure serum proteome by using two quantitatively precise proteomics assays, among the above-mentioned twin subjects. Identified protein biomarkers will be combined with metabolomics biomarkers to create a prediction model for chronic atrophic gastritis. Last, we have created a cohort of subjects who were histopathologically diagnosed with chronic atrophic gastritis or more severe precursor lesions. They were followed for stomach cancer occurrence, and a nested case-control study will be performed. Baseline formalin-fixed paraffin-embedded tissue blocks will be retrieved for both stomach cancer cases and their matched controls, and patterns of tissue proteome and transcriptome will be compared, to identify driving factors associated with progression of precursor lesions to malignancy. The results will hopefully improve our understanding of the etiological factors and provide promising early detection biomarkers for stomach cancer and its precursor lesions.

Acute myeloid leukemia (AML) is an aggressive blood cancer that is highly genetically complex and cell diversified. The heterogeneity of both tumor genetic changes and cellular subtypes means that response to treatment is very difficult to predict. The response is a crucial part of individualized cancer treatment. While the relationship between drug sensitivity and genetic changes has been largely investigated at the tissue level, studies at the single-cell level, such as cellular heterogeneity, are still limited. We will investigate how response to cancer treatment is affected by the heterogeneity of cells in AML patients. We will evaluate how drug sensitivity is associated with cell heterogeneity in AML tumors detected from single-cell data. We will investigate the diversity of changes by integrating multiple biological features including mutations, immune cell type, RNA expression and alternative splicing patterns. We will use the expanded information, other biologically meaningful features from tissue-level omics data, and advanced machine learning methods to improve prediction of response to individual cancer drugs and drug combinations for patients with AML, and to explore carcinogenic mechanisms of cancer drugs. To develop knowledge about the relationship between drug sensitivity and cell heterogeneity in AML tumors. To develop improved prediction models of drug response for patients with AML.