Development of an integrative approach for the derivation of signatures and translational analysis of cancer genomic data

The onset and constant advancement of molecular technologies has enabled the parallel, high-throughput processing of millions of sequence reads, thus ushering a new era with numerous, novel applications in basic, applied and clinical research. In this work we present an analysis framework of NGS genomic data, integrating functional and pathway analyses, for the inference of gene signatures with diagnostic and/or prognostic value. In addition, a system for the integration and processing of multi-layered data is presented, with most of the steps already operated through the Galaxy platform. In the interests of reproducibility, streamlined tools (variant calling, VCF analyzer, BioTranslator) were used, resulting in standardised workflows. A pilot analysis —part of the first national effort for the genomic characterisation of primary cutaneous melanoma— was performed including ten patients, resulting in a short list of candidate genes with a probable causative role in the disease. Ultimate aim is to develop a gene signature, which will be constantly updated through the incorporation of new cases, allowing the accurate patient classification and clustering, towards a precision medicine approach.