Cancer is a major health burden and the cause of ~50,000 deaths per year in Australia. One in four Australians will receive a cancer diagnosis during their lifetime. The introduction of immune checkpoint inhibitors (ICI) targeting CTLA-4 and PD1/PD-L1 has revolutionised the treatment of melanoma and many other solid malignancies with at times drastic improvements in clinical outcomes. Unfortunately, some patients suffer immune-related adverse events and in drastic cases, tumour hyperprogression. Subsequently, an area of critical need that has received extensive attention is the ability to predict who will benefit from immunotherapy prior to treatment. To date, only three biomarkers have been approved by the FDA for clinical use, tumour tissue PD-L1 protein immunostaining to assess the presence of the target in the tumour microenvironment (TME), tumour mutational burden (TMB) as a proxy for neoantigen load, and likewise, DNA mismatch repair (MMR) deficiency. Despite their utility, both tumour tissue PDL1 protein and TMB are limited by poor sensitivity/specificity (with responses still observed when the assay is deemed ‘negative’ and vice versa), and MMR deficiency is rare. A slight increase in the accuracy of prediction can be gained via, for example, the combined use of PD-L1 immunostaining and TMB. In a CRC-P grant awarded to Max Kelsen in collaboration with genomiQa, QIMR, and BGI, we aimed to utilise artificial intelligence and multi-omics (whole-genome DNA sequencing and whole transcriptome mRNA sequencing) from several melanoma cohorts to capture as much information about the TME as possible and utilise this information to accurately predict the response to immunotherapy.