Worldwide, ˜300,000 new cases of differentiated thyroid cancer are reported per annum and thyroid cancer now represents the most rapidly increasing cancer in the US and in the UK. In general terms, outcome is good (10-year survival >90%). However, up to 25% of patients develop local or regional recurrences, and have a significantly reduced life expectancy. We hypothesise those thyroid tumours which subsequently recur display a distinct pattern of driver mutations, and that molecular characterisation of these mutations will reveal novel mechanisms involved in thyroid tumour recurrence. Next generation sequencing was performed and whole genome sequencing data downloaded from The Cancer Genome Atlas (TCGA), prior to filtering and bioinformatic analysis. This identified mutations in a number of biologically significant genes, including Inosine-5-monophosphate dehydrogenase 2 (IMPDH2), 6-Phosphofructo-2-Kinase/Fructose-2,6-Biphosphatase 4 (PFKFB4) and Dicer 1 ribonuclease type III (DICER1), which occurred in tumours which subsequently recurred. As in silico analysis suggested all variants to be potentially pathogenic, expression vectors were obtained for each gene and site directed mutagenesis performed to recapitulate the mutations. Subcellular localisation, proliferation, cellular migration and invasion were all investigated in cell lines which represented the background driver mutation of each tumour (TPC1: RET/PTC; SW1736: BRAF; Cal62: KRas). In TPC1 cells IMPDH2 and DICER1 mutations induced significantly increased cell migration at 24 hours, and overexpression of IMPDH2 resulted in altered intracellular localisation into intracellular discrete bodies. Our on-going studies show mutated genes demonstrating some aspects of the aggressive phenotypes that would be expected to be associated with tumour recurrence. However, there are clearly different mechanisms by which these mutations are functionally pathogenic, and modelling tumour recurrence is particularly challenging. For these mutations in IMPDH2, PFKFB4 and DICER1 in silico prediction of pathogeneity is not enough to understand the complexity of tumour mutation interactions, confirming the necessity of concomitant in vitro analysis.