Figure 3: Removal of non-carcinoma cells reveals intrinsic tumour cell heterogeneity.
From: Single-cell RNA-seq enables comprehensive tumour and immune cell profiling in primary breast cancer
(a) Centred correlation matrix for all single cells demonstrates low cell-to-cell correlations (Pearson’s r) in tumours with lymph node metastases (left). After separation of carcinoma and non-carcinoma cells, cell-to-cell correlations within the same tumour group are increased (right). Each row and column represents single cells. In the colour panel on the far left side, grey represents tumour and light blue represents non-tumour cells. (b) Intratumoral correlations before (white boxes) and after (striped boxes) the removal of non-carcinoma cells (left). Each box shows the median and interquartile range (IQR 25th–75th percentiles), whiskers indicate the highest and lowest value within 1.5 times the IQR and outliers are marked as dots. Samples were ranked by mean value of cell-to-cell Pearson’s correlation coefficient (right). (c) Unsupervised PCA on the transcriptome separating patient-specific tumour groups for only tumour cells.
