high beta score values

[GeorgeGkafas]

Dear Katie,

Thank you very much for BetaScan program! It's super fast and really handy.

I've recently used the program to detect balancing selection in some SNPs of interest, knowing that they have non-synonumous substitutions.
Since, I was struggling to use an outgroup, I used the beta1 score.

I have a VCF of 60 samples of a single population. I filtered my VCF for i) remove-indels, ii) biallelic, iii) maf 0.05 and missing data up to 20% of the population.

Following your tutorial I used glactools to get the input file for each chromosome. After running the program I got some really high betascores in intergenic regions, but not for my SNPs of interest. I used several windows but still the extreme high scores are still there. The high scores are ~ 40 to 50 for a w 1000. The scores of my SNP of interest are around 4 to 5 using w 1000, values that are quite similar to your paper (and others), but when I use the suggested 1% of the whole scores, my SNPs' score of interest are way very down to the list, due to the others high scores.
When I use lower window size (e.g. 200bp) the high scores are dropped to ~7 to 8, but also my SNP's scores are also dropped (~1.8).

Also, I tried to calculate the z-score for significance, but these high scores are throwing my SNPs under the bus.

Any suggestions please for these high values? Or is it normal to get such values?

Thanks!
best,
George