Accuracy and reproducibility
In all fields of science, accuracy and reproducibility are of paramount importance and it’s never been more true than in the case of genomic sequencing where minor errors could lead to real problems further down the line for groups studying hereditary conditions or perhaps oncology. The obvious starting point would be a set of defined standards that would facilitate quantitative performance measurement and give confidence in analysing variants.
The Genome in a Bottle grants you a wish!
One of the first steps was the Genome in a Bottle Consortium established by NIST to ‘develop widely accepted reference materials and accompanying performance metrics to provide a strong scientific foundation for the development of regulations and professional standards for clinical sequencing’. The consortium first met in April 2012 and set up four groups, each tasked with coming up with potential answers.
Each group looked at…
The first group looked at selecting cell lines that would give whole genome reference material (RM) and synthetic DNA spike material to add to the RM. Others looked at the design of experiments that would sequence the RM’s using different sequencing methods, another looked at the bioinformatics and the last looked at performance metrics and figures of merit.
Growing a genome
The project is still in the initial stages and the RM’s are being considered, one of the favourable candidates was NA12878 pedigree and Personal Genome Project, and a ‘large batch’ was being grown in November 2012. In January 2013, all of the data available was being collected into one depository. There were eleven donor candidates in trios of father, mother and child. To assess the technical elements, it was agreed that some ethnic diversity would be a useful inclusion.
One cell line being investigated is not from a family group but is one that contains engineered mutations that have been added in order to test how well they are identified. Also under investigation are some DNA constructs with known cancer-associated mutations that would be useful as DNA spikes.
Patent the human genome? Not yet, but…
As an aside, a blog entry on the Genome in a Bottle Consortium, discussed the thorny question of what would happen if a patented disease gene was sequenced. Unhelpfully, there is no clear cut answer just at the moment, because there has been no definitive legal ruling. In the blog, Nicholson Price from Harvard Law School says that the US Supreme Court will be hearing a case between the Association for Molecular Pathology and Myriad Genetics, which will examine whether the human genome can be patented. In the case history so far, courts have ruled that patents on DNA were invalid, but that ruling was later overturned. The Supreme Court should provide some answers during 2013.