Extended Data Fig. 3: Prognostic value of micrometastasis- associated genes in basal-like breast cancer patients. | Nature Cell Biology

Extended Data Fig. 3: Prognostic value of micrometastasis- associated genes in basal-like breast cancer patients.

From: Transcriptional diversity and bioenergetic shift in human breast cancer metastasis revealed by single-cell RNA sequencing

Extended Data Fig. 3

a, Kaplan-Meier curves show relapse free survival (RFS) in basal-like breast cancer patients from the KM plotter database (879 patients), based on their primary tumour expression of specified micrometastasis-associated genes. P values were determined via a log-rank test. b, Schematic for the construction of a stepwise logistic regression model to identify top biomarker candidates descriptive of primary tumour or micrometastatic cells. The data was subsampled to analyse equal numbers of micrometastatic and tumour cells from each mouse. The model was run on 10 subsamplings of the data, with the number of genes in each model determined by AIC. c, Plot demonstrating the AIC versus the number of genes included in the model. AIC is used to balance parameter additions (that is gene additions) with the descriptive power of a model. Data is presented as the 10% and 90% quantiles of the 10 data subsamplings. d, Bar plot showing the number of model appearances for each gene out of 10 data subsamplings.

Back to article page