I am using plink to ld-pruning SNPs. But I wonder why plink always produce the same pruning results, even when I add a random seed.

Let's suppose, SNPs A and B are in LD. They can be also in LD with other SNPs. The SNPs in LD with A and the SNPs with B might be different. Thus, the first choice of A or B might produce different pruning SNP set. So, LD-pruning should produce a random result, really?

Here, I worry that Plink always produces a same result, it will produce a bias? Thank you for comments.