The study aims to comprehensively characterize the genomic and transcriptomic profiles of HER2 positive breast cancer (BC) and their evolution under the pressure of modern targeted therapies. Here, we will integrate multi-omics data from 3 methods including whole exome sequencing (WES), CUTseq, and RNA sequencing (RNA-seq) to investigate 197 HER2-positive BC with ~500 tumor samples longitudinally collected from the randomized phase 2 PREDIX HER2 trial (ClinicalTrials.gov Identifier: NCT02568839). The association between recurrent somatic variants and pathologic complete response (pCR) rate will be assessed for standard combination of docetaxel, trastuzumab, and pertuzumab or trastuzumab emtansine (TDM-1) monotherapy, respectively. Likewise, other genomic profiling such as mutation signatures, tumor mutation burden, and copy number scores as well as transcriptomic characteristics like intrinsic subtypes, pathway scores and immune cell deconvolution will be calculated, whose predictive values on targeted therapies will be evaluated.