Genotyped Data Vs Imputed Data

[Dovini Jayasinghe]

Hi,

I'm new to statistical genetics and hence the PRSice software. I'm hunting for basic meanings/definitions related to the context. I want to pinpoint clear differences between "Genotyped Data" and "Impute Data" where I could not find exactly the same terms. It is known that PRSice handles both "genotyped" and "imputed" data (Source: doi: 10.1093/gigascience/giz082), but what are they really?

 Cheers!

[Sam Choi]

To save cost, most of the time we only genotype part of the genome. What we can do is then perform something we call imputation, using the LD structures between variants to "guess" what the variant should be in area that we did not genotype directly. This guesses ain't always correct, and have uncertainties associated with that, so we might want to take that into consideration. You can search for "genotype imputation" on google or biostar for more info.

PRS analyses are usually considered as secondary analyses, so it is advice that one should be familiar with the basic of genetics analyses before starting PRS as that might sometimes be confusing, especially we usually write our documents as if users have basic knowledge on all primary genetic analyses.

[Dovini Jayasinghe]

Thanks Sam