Skip to main content

myCMIE 2.0

15 Expansion of Exploration

15.1 CCLE-Specific Features

15.1.1 Drug Sensitivity Oncoplots

  • Input Methods: Query drug sensitivity via IC50 values from the CREAMMIST database, using either “Upload” for batch compound names or “Typing” for real-time direct input.

  • IC50 Values Display: By default, oncoplots show log2-transformed concentration (µM). Normalization options within compounds are available to facilitate cross-sample drug sensitivity comparison.

15.1.2 Pathway Oncoplots

These plots utilize cell line-specific driver genes from the GDS database for enrichment analysis. Users can select reference gene sets and cluster samples by enrichment results to explore gene set enrichment prevalence and match status.

15.2 TCGA-Specific Feature

15.2.1 Immune Abundance Oncoplots

Derived from TCGA and CCLE datasets, these plots display immune abundance for each sample or cell line, with data from CIBERSORT (RNA-seq based) and methCIBERSORT (methylation-based) for TCGA and CCLE.

15.3 PDX Specific Feature

The PDX (Patient-Derived Xenograft) model data utilized in this module is sourced from the PDMR (Patient-derived Model Repository) and BCM (Baylor College of Medicine), focusing on breast cancer. These models facilitate the visualization of drug response data, which is annotated with various responses such as Progressive Disease (PD), Partial Response (PR), Stable Disease (SD), among others.

Input Flexibility: Users are provided with a dynamic and responsive interface to select and examine compounds of interest. Given the diversity of compounds tested across different PDX models, the selection list adapts accordingly. Only compounds that have been tested on the currently selected PDX model are displayed, streamlining the user experience and preventing data overload for a more efficient and user-friendly interface.

15.4 Common Oncoplot Features

15.4.1 Expression Oncoplots

Expression oncoplots enable a comprehensive exploration of RNA expression alongside mutation and copy number data. Supported by TCGA and CCLE databases, this feature draws from the cbioPortal API, with a z-score normalization applied to diploid samples. Here, a z-score above 0 denotes “high” expression, while a z-score of 0 or below indicates “low” expression for each gene.

15.4.1.1 Gene Selection and Input.

Users can introduce a gene list from the patient profile or input a custom list. Alternatively, both lists can be merged for an integrated analysis. The visualization can be tailored by selecting specific genes in the ‘Select Genes’ section for display.

15.4.1.2 Expression Analysis.

Column-wise clustering is available to categorize genes with similar expression profiles across the samples or cell lines. Clusters are analyzed to summarize the overall expression pattern, identifying whether a cluster exhibits high expression in any sample.

15.4.2 Survival Oncoplots Integration

This module graphically represents survival status and time, visualizing the data as color-coded heatmaps. Available data spans TCGA, CCLE, MSKCC, and METABRIC databases. The survival duration, denoted in months, is illustrated, with an asterisk (*) marking deceased patients.

15.4.3 Non-B Burden Assessment

The application includes our proprietary non-B related burdens calculation, as defined by Xu and Kowalski (2022):

  • nbTMBp: This metric represents the percentage of non-B related mutations relative to the total Tumor Mutation Burden (TMB) for each sample.

  • mlTNB: Mutation-localized non-B Burden assesses non-B forming sequences situated at mutation sites within tumor samples, quantified by non-B types.

Users can normalize these burden metrics across samples (column-wise) and may incorporate IC50 oncoplot data to investigate the correlation between non-B related burdens and drug sensitivity.