Abstract Backgound and Aim: The best test to guide the choice of systemic therapy for breast cancer (BC) has not yet been identified. We did this study to identify factors that drive proliferation features in BC and assess their association with clinical outcomes after systemic therapy. Methods: We applied an artificial neural network-based integrative data mining approach to three cohorts of patients with untreated lymph node (LN)-negative BC (Wang et al; n=286, Desmedt et al; n=198 and Schmidt et al; n=200). The results were validated in four cohorts of BC patients (the Nottingham discovery cohort (n=171), Uppsala cohort (n=249), The Cancer Genome Atlas-Breast Cancer project [TCGA-BRCA; n= 970] and Molecular Taxonomy of Breast Cancer International Consortium [METABRIC cohort; n=1980]. Genes that featured prominently in our interactome map of proliferation have been chosen to take them forward to investigate their clinicopathological relevance of their gene copy number aberrations (CNAs), mRNA transcript expression, and protein expression and their associations with breast cancer-specific survival (BCSS), distant relapse-free survival (DRFS) and pathological complete response (pCR) in ten international cohorts of BC (n>12000 patients). Findings: ESR1, SPAG5, EGFR, BCL2, and FOXA1 were among the 39 common gene probes that were predictive across most proliferation features and datasets. In TCGA-BRCA cohort, SPAG5 gene mutation, gain/amplification and loss at the Ch17q11.2 locus were detected in 43 (4.4%), 177 (18.2%) and 180 (18.8%) of 970 patients, respectively and 65 (31%) of 479 ER-positive /HER-positive patients showed gain/amplification of SPAG5 gene. In multivariable analysis, high SPAG5 transcript and SPAG5 protein expression were associated with reduced BCSS compared with lower expression (METABRIC: HR 1·27, 95% CI 1·02–1·58, p=0·034; untreated LN-negative cohort: 2·34, 1·24–4·42, p=0·0090; and Nottingham-cohort: 1·73, 1·23–2·46, p=0·0020). In patients with ER-negative/HER2-negative or ER-positive/HER2-negative BC, high SPAG5 transcript expression was associated with an increased pCR compared with low SPAG5 transcript expression after receiving anthracycline neoadjuvant chemotherapy (AC-NeoACT) [(Multicentre phase 2 clinical trial cohort; n=136; OR 2·47, 95% CI 1·17–5·21, p=0.016) and (MD Anderson- taxane+AC-NeoACT cohort; n=287; OR 3·16, 95% CI 1·46–6·84, p=0.003); respectively]. In patients with ER-positive/HER2-negative BC who received taxane+AC-NeoACT followed by adjuvant tamoxifen (Adj-Tam) for 5 years (MD Anderson- taxane+AC-NeoACT cohort; n=287), high and low SPAG5 transcript expression had similar DRFS (HR 1·40, 95% CI 0.76–2·58, p=0.282). Whereas in ER-positive/HER2-negative BC patients who received only adj-Tam (n=298), high SPAG5 transcript expression was associated with reduced DRF at 5 years compared with lower expression (HR 1.98, 95% CI 1.19–3.27, p=0.008). Interpretation: The transcript and protein products of SPAG5 are independent prognostic and predictive biomarkers that might have clinical utility as biomarkers for combination cytotoxic chemotherapy sensitivity in ER-positive/HER-negative BC. Citation Format: Abdel-Fatah TMA, Agarwal D, Zafeiris D, Pongor L, Györffy B, Rueda OM, Moseley PM, Green AR, Liu D-X, Pockley AG, Rees RC, Caldas C, Ellis IO, Ball GR, Chan SYT. Identification of proliferation related derivers and their roles in precision medicine for breast cancers: A retrospective multidimensional comparative, integrated genomic, transcriptomic, and protein analysis [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P6-09-16.