Synthetic lethality describes a pair of gene functions where losing either one alone is survivable, but losing both together kills the cell. The classic clinical example is the pairing of BRCA1 or BRCA2 loss with inhibition of poly(ADP-ribose) polymerase, or PARP, an enzyme that helps repair single-strand DNA breaks. A BRCA-mutant tumor cell has already lost homologous recombination, its high-fidelity route for repairing double-strand breaks. When PARP is inhibited, unrepaired single-strand breaks collapse into double-strand breaks at replication forks, and the tumor cell has no reliable way to fix them. A normal cell retains at least one functional BRCA allele, so it repairs those breaks by homologous recombination and survives.
This is selectivity that does not depend on proliferation rate. It depends on a tumor-specific molecular defect, which is why it is a genuine extension of the selectivity logic rather than a variation on it: instead of exploiting how fast a cell divides, we exploit what the cell has already lost. The simulation lets you vary the degree of PARP inhibition and the BRCA status of the cell population, and watch how survival separates between the two genotypes.