CMScaller

CMScaller classifies colorectal cancer (CRC) samples into consensus molecular subtypes (CMS) using cancer cell-intrinsic gene expression markers to enable subtype identification across pre-clinical models.

Key Features:

• Subtype Classification: Classifies CRC into four biologically defined consensus molecular subtypes (CMS).
• Cancer cell-intrinsic marker filtering: Utilizes a filtered set of cancer cell-intrinsic, subtype-enriched gene expression markers to reduce signals from immune and stromal compartments and improve identification of subtypes including the CMS4-mesenchymal group.
• Application across pre-clinical models: Recapitulates CMS subtypes in immortalized cell lines, patient-derived organoids, and xenografts, validated across 551 samples.
• Predictive drug-response associations: Integrates public drug response data from patient-derived xenografts and cell lines to identify CMS–drug associations and predict responses to standard CRC treatments.

Scientific Applications:

• Translational subtype mapping: Maps clinical CMS to pre-clinical models to enable translational studies.
• Stratified therapeutic evaluation: Supports identification of CMS-specific drug responses and CMS–drug associations for stratified therapeutic strategies.
• Prognostic subgroup identification: Enables identification of the poor-prognostic CMS4-mesenchymal subgroup based on cancer cell-intrinsic expression.
• Biomarker and mechanism studies: Provides cancer cell-intrinsic, subtype-enriched gene expression markers for biomarker discovery and mechanistic investigation.

Methodology:

Uses a filtered set of cancer cell-intrinsic, subtype-enriched gene expression markers to classify samples into four CMS categories; integrates public drug response data from patient-derived xenografts and cell lines; validated on a dataset of 551 samples comprising immortalized cell lines, patient-derived organoids, and xenografts.