Tumor Type

Skin Cancer - Melanoma
Project Profile
Funding Organizations
Research Organizations
Research Activities
Publication Policy

1:   ICGC Goals, Structure, Policies and Guidelines Section E.3 - Publication Policy HTML 
2:   Template Letters to Facilitate Communications HTML 
to ensure appropriate dialogue between data users and generators and for authors to contact ICGC members and editors
Clinic & Pathology

Australia: Melanoma Institute Australia
Royal Prince Alfred Hospital Tissue Pathology and Diagnostic Oncology
QIMR Berghofer Medical Research Institute
Peter MacCallum Cancer Centre
Olivia Newton-John Cancer Research Institute
The University of Sydney

Sample preparation and clinical data:
All genomes sequenced as part of the AMGP are derived from well-characterised clinical material which is accurately annotated with clinico-pathological, treatment and outcome data. All samples were pathologically assessed prior to inclusion into the study, with samples requiring greater than 80% tumour content and less than 30% necrosis to be included. Acral melanomas were classified as occurring within acral skin of the palm of the hand, sole of the foot and under nail beds. The lack of hair follicles, thickened stratum corneum and clinical site was confirmed in all cases. Mucosal melanomas were defined as occurring in the mucosal membranes lining oral, respiratory, gastrointestinal and urogenital tracts.

Sequencing & Analysis

Australia: Melanoma Institute Australia
QIMR Berghofer Medical Research Institute

Whole-genome sequencing was performed on either Illumina Hiseq 2000 sequencer at 100 bp pair-end libraries prepared using Truseq SBS V3-HS kits kits (Illumina) or Illumina Hiseq X ten sequencer at 150bp PE using the Truseq Nano DNA kit.

Sequence data was aligned to the GRCh37 assembly using multi-threaded BWA 0.6.2-mt resulting in sorted lane level files in sequence alignment/mapping (SAM) format which were compressed and converted to binary file (BAM) created by samtools 0.1.19. Sample-level merged BAMs, one each for matched germline and tumour samples were produced by in-house tools and duplicate reads marked with Picard MarkDuplicates 1.97 (http://picard.sourceforge.net). Quality assessment and coverage estimation was carried out by qProfiler and qCoverage (http://sourceforge.net/projects/adamajava). To test for the presence of sample or data swaps all sequence data were assessed for concordance at approximately 1.4 million polymorphic genomic positions including the genotyping array data by qSignature.

Somatic mutations and germline variants were detected using a dual calling strategy using qSNP and GATK and indels of 1–50 bp in length were called with Pindel and GATK. All mutations were submitted to the International Cancer Genome Consortium (ICGC) Data Coordination Centre (DCC). Mutations were annotated with gene consequence using Ensembl gene annotation with SnpEff. Somatic genes which were significantly mutated were identified using two approaches: MutSigCV and OncodriveFML 1.1 using a threshold of q<0.1. Significant non-coding elements were detected using OncodriveFML 1.1. Somatic copy number and ploidy were determined using the TITAN tool. SV were identified using the qSV tool and chromosomes containing highly significant non-random distributions of breakpoints were identified.

Complementary Studies

Details to follow

Data Storage, Analysis & Management

Australia: QIMR Berghofer Medical Research Institute
The University of Sydney

Somatic variants for this study have been submitted to the ICGC Data Coordination Centre (DCC) and are publicly available through the ICGC DCC Data Portal (https://dcc.icgc.org). The BAM files for seventy of the donors in this study have been submitted to the ICGC PanCancer Analysis of Whole Genomes (PCAWG) and these will be transferred to the European Genome/Phenome Archive (EGA) by PCAWG. Data from the remainder of the donors will be transferred to the EGA by the analysis team. The EGA accession for this study is EGAS00001001552.

Project Summary

The Australian Melanoma Genome Project (AMGP) is performing high-depth whole genome sequencing to determine genome-wide patterns of mutation in over 400 melanoma patients. In Australia, melanoma is the third most common cancer in both men and women, with an overall lifetime risk of 5.8%, the highest incidence rate worldwide. Melanoma is molecularly diverse and displays one of the highest mutation loads of any cancer, which has been attributed to ultraviolet radiation exposure, however, melanoma can arise in non-UVR exposed sites of the body. Therefore, the AMGP is sequencing a range of melanoma histology subtypes, including a large cohort of cutaneous melanomas and melanomas that have arisen in mucosal surfaces, the uveal tract of the eye, and on the hands and feet; the latter (acral) sites account for a much higher proportion of melanomas in Asians, than in Europeans. Elucidating the significant mutational processes and genomic drivers in all types of melanoma is important for its epidemiology and prevention and for identifying new therapeutic leads required for melanoma treatment globally.

Principal Investigators

Melanoma Institute Australia
• Graham Mann – Co-lead
• Richard Scolyer – Co-lead
• James Wilmott – Project manager

QIMR Berghofer Medical Research Institute
• Nick Hayward – Co-lead
• Peter Johansson
• John V. Pearson
• Nicola Waddell

Lead Jurisdiction