Increasing the number of intersecting SNPs

[Boga]

Hey, I'm new to PRSice, so I hope this isn't a stupid question.

I'd like to calculate PRS. I have base data that only contain rather few SNPs (20), of which only 10 are present in my target data (already imputed). The base data are from a published GWAS, and unfortunately I don't have access to the full data to prune them differently.

Are there any reasonable methods to increase the intersect? I thought maybe I could look on LDlink's LDproxy tool, and manually check if any SNPs have a decent proxy (e.g. R² > 0.7). If so, I was thinking I could change maybe swap out those SNPs in the base data, while keeping the effect size and P-value ..

Is something like this reasonable to use for PRSice, or are there any proper solutions to this?