MOSAIC: Biospecimen Core

SUMMARY: BIOSPECIMEN CORE The major objective of the Biospecimen Core is to obtain, process, image, and analyze glioblastoma (GBM) patients’ biopsies for use in both Projects of our Mayo/Columbia CSBC Mathematical Oncology Systems Analysis Imaging Center (MOSAIC). Patient biopsies are central to our CSBC’s primary goal: building a conceptual framework to understand sensitive response to therapy through the represented tissue states in clinical and experimental settings. The duties of the Biospecimen Core are spread across both sites: Mayo Clinic Arizona (MCA) and Columbia University Medical Center (CUMC). The Biospecimen Core will assist the Center by operating under two specific aims. In Aim 1, we collect, process, and analyze image-guided biopsies from pre-and post-treatment GBM. For Aim 2, we perform histopathological, immunohistochemical (multiplex-IHC), and molecular analyses (bulk RNA-seq, scRNA-seq) of image-guided biopsies. The Biospecimen Core will provide support for both projects. Project 1: Targeting Glioma Tissue States will utilize the RNAseq and single cell sequencing data from the biopsies collected and analyzed by the Biospecimen Core to assess the compositional landscape of glioblastoma along with the cross-talk between different subgroups that influence phenotype. Project 2: Imaging the Dynamic Tissue State in Patients In Vivo will utilize the bulk histology and RNAseq along with the imaging of the image-localized biopsies collected from the Biospecimen Core. Project 2 will also incorporate these histologic and molecular cell lineage data into ML and mechanistic models that map the relationship of magnetic resonance imaging (MRI) features to cellular composition. These data will drive the training and validation of machine learning models and the parameterization of mechanistic mathematical models. The data generated by the Biospecimen Core will be transferred to the Data Integration and Computation Core, where it will be stored, further processed including batch correction and atlas alignment, and integrated into computational modeling tools.