What does this equation mean?

It is worth working through what that channel can and cannot do to a population’s allele frequencies, because the arithmetic is genuinely calculable and genuinely modest. Consider a recessive disease allele at population frequency q , and suppose that in some fraction of pregnancies where both parents are identified carriers, embryo testing prevents the birth of an affected (homozygous) child while parents still have the same number of children overall, drawn from the surviving unaffected and carrier embryos. The standard discrete-generation recursion for selection with coefficient s against a fully recessive genotype is: $q_{n+1} = \frac{q_n(1 - s q_n)}{1 - s q_n^2}$. Set s = 1 , the theoretical maximum in which every affected embryo is detected and never implanted: for q = 0.01 , this gives $q_{n+1}$ = $q_n$/(1+$q_n$) $\approx$ 0.009901 , a reduction of under one-tenth of one percent of the allele’s frequency in a single generation, even at total, universal, perfectly effective screening. The reason the effect is so small is structural, not a modeling artifact: for a rare recessive allele, the overwhelming majority of copies are carried silently in heterozygous carriers, invisible to any selection that acts only on homozygotes, and embryo selection — like natural selection against recessive disease before it — can only ever act on the small fraction of matings where two carriers happen to pair up. Layered on top of that structural limit is the fact that PGT operates only within IVF, itself a minority pathway to conception worldwide, so the realized value of the effective screening fraction f across all human conception is far below the illustrative s=1 ceiling used above. Screening’s population-genetic footprint, in other words, is real, directionally opposite to the tiny drift toward higher mutation load discussed earlier, and — on the arithmetic above — slower than either.