Skip to main content
Fig. 1 | Genome Medicine

Fig. 1

From: Beyondcell: targeting cancer therapeutic heterogeneity in single-cell RNA-seq data

Fig. 1

The Beyondcell workflow and Beyondcell switch point. Beyondcell is a methodology for the identification of drug vulnerabilities in scRNA-seq data. Using Beyondcell, we have identified the presence of therapeutic clusters in our data, defined as sets of cells sharing a common behaviour towards a collection of drugs. a Given two inputs, an scRNA-seq expression matrix and a drug signature collection—either the drug perturbation (PSC) or the drug sensitivity (SSC) collections or a user-provided GMT file/ranked matrix—Beyondcell calculates a score (BCS) for each drug-cell pair. The resulting BCS matrix is used to determine the presence of therapeutic clusters, which can be visualised using a UMAP in Beyondcell. A sensitivity-based ranking can be obtained in order to prioritise the best hits. b The scaled BCS ranges from 0 to 1 and measures the cell perturbation susceptibility (when using the PSC) or the predicted sensitivity to a given drug (when using the SSC). The BCS can also be used to evaluate the cells’ functional status if functional signatures are applied. Furthermore, a switch point (SP) is calculated for each analysed signature, by determining the value in the 0 to 1 scale where cells switch from a down-regulated status to an up-regulated one. Thus, the most therapeutically-homogeneous tumours would be those in which each and every one of their cells responds with the same directionality to a certain drug, either towards sensitivity (SP = 0) or resistance (SP = 1), while a heterogeneous response would be represented by intermediate SPs

Back to article page