Genechip human 1.1 st arrays

Gene expression microarrays were performed at the Genomics platform of Institut Curie. GeneChip Human 1.1 ST arrays were hybridised according to Affymetrix recommendations, using the Ambion WT Expression Kit protocol (Life Technologies) and Affymetrix labelling and hybridisation kits. Affymetrix CEL files were imported into the Gene Expression Workflow in Partek^® Genomics Suite version 7.0 (Partek Inc., St. Louis, MO, USA, www.partek.com). Background correction, quantile normalisation, log2 transformation, and probeset annotation were performed using default settings for the Robust Multichip Average (RMA) procedure.
Gene set enrichment analysis (GSEA, v4.0.3) software and MsigDB database (v7.0)^{44 (link)} were used to identify overrepresented biological functions for differentially expressed genes between PDX and primary breast tumours.
Enriched pathways were then represented as an interactive network, using EnrichmentMap^{45 (link)}, a Cytoscape^{46 (link)} application that determine relationships between pathways. Pathways are represented by nodes and connected by edge if they shared common genes. Highly interconnected nodes are clustered in order to identify major biological processes. Affymetrix CEL files and normalised log2 RMA data are available at the GEO database (accession No. GSE146661; [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE146661]).

Transcriptomic profiling of 39 TNBC PDX was performed by gene expression arrays as previously described [14 (link)]. GeneChip Human 1.1 ST arrays were hybridized according to Affymetrix recommendations, using the Ambion WT Expression Kit protocol (Life Technologies) and Affymetrix labeling and hybridization kits. Affymetrix CEL files were imported into the Gene Expression Workflow in Partek Genomics Suite version 7.0 (Partek Inc., www.partek.com). Background correction, quantile normalization, log2 transformation, and probeset annotation were performed using default settings for the Robust Multichip Average (RMA) procedure. The molecular subtypes of TNBC PDX models were determined with the web‐based subtyping tool TNBCtype [14 (link), 18 (link)]. Given a gene expression data matrix, this tool displays for each tumor sample the predicted subtype, the corresponding correlation coefficient, and the permutation P‐value [18 (link)].

For gene expression microarray analyses, GeneChip Human 1.1 ST arrays were hybridised according to Affymetrix recommendations, using the Ambion WT Expression Kit protocol (Life Technologies) and Affymetrix labelling and hybridisation kits, as previously described in ref. ^{13 (link)}. Gene set enrichment analysis (GSEA) software and MsigDB database^{55 (link)} were used to identify overrepresented biological functions for differentially expressed genes between PDX and primary breast tumours.