CARMA: Copy Aberration Regional Mapping Analysis. A main theme of the group is DNA analyses and in particular to decipher DNA copy-number alterations (CNAs). In this project we recently published a framework with several mathematical defined indices that recognize different types of CNAs in breast cancer. We find it to be of clinical relevance, and we now pursue to integrate the CARMA framework with DNA mutation data as it strengthens the ability to reveal evolutionary patterns, aiming to investigate whether or not this can predict patients at risk for none, early or late relapse. In a previous project: Liquid biopsies – optimizing methodologies for nucleic acid and single cell detection, capture and analysis, we have developed robust methods for ctDNA analyses (partly through the EU/IMI collaboration CANCER-ID). We are now analyzing the CNAs in tumor samples from 200 patients taken at four time-points during neo-adjuvant treatment (I-BCT trial, PI: O. Engebråten/O. C. Lingjærde) by WGS and RNAseq and will follow with analyses of ctDNA from the same patients. This will reveal both different evolutionary patterns but also predictive biomarkers of prognosis and therapy response. The CANCAN (CANcer specific Copynumber alteration Analysis) project is related to CARMA but here we aim at fitting the algorithms to targeted sequencing data suited for diagnostic use. We have customized a targeted NGS panel and can now perform allele-specific copy number assessment. The test is at present being validated in a clinical cohort. TREAC: Towards personalized treatment for patients with aggressive breast cancer.This is a recently initiated work with Engebråten/Naume/Mortensen/Valla where we are in particular interested in estrogen receptor positive breast cancer patients with aggressive disease. We are now performing whole genome and RNA sequencing from the clinical trial EMIT and will apply the CARMA algorithms to these selected samples. The aim is to identify molecular alterations to identify the patients but also to reveal potentially targetable targets, and finally to build a diagnostic test based on the CANCAN set-up.