Longitudinal Multi-Omic Characterization of Patient Responses to Neoadjuvant HER2-Targeted Therapy

This dataset contains longitudinal multi-omic profiles derived from patients with early HER2-positive breast cancer enrolled in the randomized phase II PREDIX HER2 clinical trial (NCT02568839). The study compared neoadjuvant treatment with docetaxel, trastuzumab, and pertuzumab (DHP) versus trastuzumab emtansine (T-DM1). The primary aim of this data collection was to monitor tumor evolution and tumor microenvironment dynamics during HER2-targeted therapy. Data Material and Collection: The dataset comprises multi-omic data from patient samples collected at different timepoint: on-treatment (after two cycles of neoadjuvant therapy) and post-treatment (after six cycles of neoadjuvant therapy). Methodology and Assay: Comprehensive molecular characterization was performed using multiple high-throughput technologies. This includes whole-exome sequencing (WES) to identify somatic mutations and genomic alterations, RNA-sequencing (RNA-seq) for bulk transcriptomic profiling, and proteomics to monitor protein abundance. Additionally, imaging-based spatial transcriptomics was performed using the Xenium In Situ platform to map the spatial organization of the tumor microenvironment; note that the Xenium spatial data is exclusively available for the on-treatment time point. The dataset consists of the following: 1) RNA-seq/sequencing_batch/Sample.fastq.gz Folder containing raw paired-end FASTQ files from RNA-seq. Each tumor sample was multiplexed across sequencing sample/lane index combinations. File names begin with the project-sample ID and include R1 or R2, denoting forward and reverse reads, respectively. 2) WES/sequencing_batch/Sample.fastq.gz Folder containing raw paired-end FASTQ files from RNA-seq. File names begin with the project-sample ID and include R1 or R2, denoting forward and reverse reads, respectively. 3) Xenium_5k Each subfolder contains the standardized, uncompressed output directory generated by the Xenium Ranger analysis pipeline for an individual tissue section. The entire dataset is over 10,000 files and over 10TB in size.