Patients

Tissues were identified from melanomas diagnosed at the University Hospitals of Leicester. Eligibility criteria were: primary cutaneous melanoma tissue, diagnosis 2004-2011, invasive disease and resident in Leicestershire. The flow of patients through the study is shown in Figure S1. Samples were assessed sequentially by pathology accession number, thus representing an unbiased retrospective cohort. For patients with multiple primary melanomas, the highest stage melanoma was included or, if they were the same stage, the first diagnosed. From a starting population of 1104, a total of 134 (12%) Leicester cases were excluded to yield 970 cases. Excluded cases are shown in Table S1. A set of 359 primary melanomas diagnosed in 2003-2005 and 2008-2010 from Nottingham University Hospitals was provided with anonymised data allowing determination of stage and melanoma specific survival. This study was granted a favourable opinion from an NHS research ethics committees in Leicester and Nottingham. The Leicester and Nottingham cases are henceforth referred to as training and validation cases respectively.