Search

Additional file 1. Fig. S1: t-distributed stochastic neighbor embedding (t-SNE) plots showing the estimated tumor purity of The Cancer Genome Atlas samples based on manual estimation using light microscopy (a), RNAseq data (ESTIMATE; b), copy number variations (ABSOLUTE; c), DNA methylation (LUMP; d) and a combined score (CPE; e).

Additional File 4: Fig. S4. Concordance between wtRNAseq and targeted RNAseq of gene expression levels and transcriptional signature scores. Distribution of Pearson correlation (Rho), bias, scale and location coefficient (components of concordance correlation coefficient (CCC)) for all 28 targeted genes within each protocol. The green points represent the two transcriptional signatures.

Additional file 2. Fig. S2: Flowchart showing the composition of the study cohort with samples from the publicly available “The Cancer Genome Atlas” (TCGA) dataset as well as specimens from the clinical GeparSixto and GeparSepto trial. Abbreviations: BRCA = breast cancer; ER = estrogen receptor; PR = progesterone receptor; TNBC = triple negative breast cancer; t-SNE = t-distributed stochastic neighbor embedding.

Additional file 1: Fig. S1. Technical variance of expression measurements in targeted RNAseq data using FF samples. (A) Distribution of absolute difference in expression between replicates for 28 targeted genes within each protocol. (B) Distribution of absolute difference in molecular signature score between replicates for 2 signatures within each protocol.

Additional File 3: Fig. S3. Distribution of raw read depth across all gene targets in FFPE samples. Line type (solid or dashed) distinguishes two technical replicates, while color represents different protocol. Genes are sorted by average read depth in all samples.

Additional File 2: Fig. S2. Distribution of raw read depth across all gene targets in FF samples. Line type (solid or dashed) distinguishes two technical replicates (6 samples have only 1 replicate in FF), while color represents different protocol. Genes are sorted by average read depth in all samples.