In the “Polygenic Risk Score(s)” section of the CanRisk tool, it is possible to calculate the PRS by either uploading a VCF datafile, or by entering the PRS value and an alpha value, for breast and/or ovarian cancer.
If entering a value, the PRS value is should be entered as a normalised z-score. A raw PRS can be converted to a z-score using the mean (mu) and standard deviation (sigma) of the PRS using the formula:
z-score = (rawPRS - mu) / sigma
mu and sigma refer to the mean and standard deviation of the PRS in the population. These parameters can vary by population, and you should take care to use appropriate values for the population relevant to the consultand. For the standard PRS, these have been estimated in Europeans in the BCAC dataset as:
BCAC315: mu=-0.424, sigma=0.611
BRIDGES306: mu=-0.422, sigma=0.608
PERSPECTIVE295: mu=-0.448, sigma=0.60
see What variants are used in the PRS?
alpha - the square root of the proportion of the overall BOADICEA polygenic variance explained by the PRS. A real number between 0 and 1.
The alpha parameter for
standardseveral PRS implemented on CanRisk, can be set by selecting a reference file from the drop down menu. When alpha and
betaz-score have been added a graphical representation of the polygenic risk is displayed.
WARNING: Selecting 'OTHER' from the reference list means you will be using a polygenic risk score distribution that is not recognised as one that is already implemented in the risk models. It is still possible to use CanRisk with the alternative PRS, but it is important to ensure that the parameters of the alternative PRS
isare correctly specified. Please refer to 'Incorporating alternative Polygenic Risk Scores into the BOADICEA breast cancer risk prediction model' for further details on defining the alternative PRS parameters".
