Breast cancer is the major malignant disease in women and has a continuously rising incidence. It is a remarkably heterogeneous disease in respect of genetic predisposition, tumor biology and clinical course. Better biomarkers are needed to improve diagnosis, prognostication and treatment prediction, to allow use of more precise drugs and avoid overtreatment. New sequencing technology provides thorough characterization of the genome and transcriptome. Breast tumors can be classified into subtypes based on patterns of gene expression, DNA methylation, nucleotide substitutions and genomic rearrangements. Pathogenic germline sequence variants in two major susceptibility genes BRCA1 and BRCA2 confer a high relative risk and explain a proportion of familial breast cancer. Few other high-risk genes are known and current knowledge supports a polygenic model, a role of common low-risk variants that may interact in multiplicative fashion, but also of rare intermediate-risk gene variants. Modern medicine relies on successful translation of new molecular research findings into the clinic, a critical step that is facilitated by forming networks and close interactions between scientists and clinicians.