myVision™ employs a number of classification techniques, some of which have been used by geneticists for many years. Myriad also uses a number of novel, proprietary techniques made possible by Myriad’s ability to test many patients. We are continually working towards creating and perfecting new methodologies to help classify variants.
- Mutation co-occurrence
- Literature evaluation
- Segregation analysis
- In trans co-occurrence
- Population frequency
- Protein structural analysis
- In silico splicing analysis
This family history-weighting tool, unique to Myriad, compares the severity of family histories of patients who carry a specific variant to that of families who carry known deleterious mutations and to families in whom no mutation was detected.
This statistical model is unique to Myriad and is used to discover benign variants, which leverages the rarity of multiple deleterious mutations co-occurring in a single patient.
Myriad’s scientists and statisticians continually evaluate the published literature, a process that often includes a reanalysis of experimental data to determine whether the data is clinically significant.
The classic gold-standard approach to variant classification where variants are tracked in families to determine if they are associated with cancer.
In trans co-occurrence
For some genes, deleterious mutations on both copies of the gene can cause well-defined and easily distinguishable diseases. Thus, variants co-occurring with known deleterious mutations on opposing gene copies in patients absent the relevant disease are more likely to be benign.
Protein structural analysis
In silico splicing analysis
While it is standard laboratory procedure to evaluate novel variants for potential RNA splicing effects using publicly available in silico methods, Myriad also uses a proprietary in silico tool for this evaluation.