Interestingly, as the melanoma

DNA yield decreased, there Bafilomycin A1 in vivo was little drop-off in the percentage of BRAF or NRAS mutations detected using either ARMS or sequencing. This would suggest that even at low DNA assay input the samples were representative of the tumours and that at low DNA input there were probably few, if any, false negative results. GSK872 analysing all samples was a good strategy to maximise the numbers of mutations detected in this study set where 88% of the samples yielded detectable DNA. In a research setting one of the strengths of sequencing is that it detects unknown mutations as well as known ones. However, in a clinical setting it is likely that decisions will be made on the basis of known characterised mutations. When analysing genes LY2874455 ic50 where mutations are found clustered in one or two exons, like KRAS, much less DNA is required for sequencing than for ARMS, although this can be reduced by multiplexing ARMS reactions. This can be an advantage when only very

small biopsies with low DNA are available. Sequencing also offers an advantage when genes contain many mutations throughout the coding region, such as p53, BRCA and APC. To develop the potentially hundreds of individual mutation detection assays required would be extremely time-consuming and require positive mutation controls to show next that the assays are functioning correctly. Sequencing reactions tend to be easier to develop and standard genomic DNA is an adequate control. It was important when performing sequencing that at least two independent PCRs were

performed from the original genomic DNA to eliminate false positive errors. We were able to distinguish true mutations from artefactual mutations by only accepting mutations detected in at least two amplicons in forward and reverse sequencing directions. Approximately 2% of the exons sequenced contained an artefact. These were most commonly detected in samples with low DNA, probably because they were not masked by more abundant unaltered DNA. These artefacts are presumably caused by damage to the DNA during fixation in formalin. None of the artefacts found in singleton were known mutations. They were not reproducible in any subsequent PCR from the original DNA samples and we were unable to validate them using other mutation discovery methods including denaturing high-performance liquid chromatography, and cloning and sequencing. ARMS appeared to be less affected by DNA artefacts as the assays only targeted known mutations. Pathology information was also taken into account as this could often explain why mutations were present at a low level in a sample.