Lentiviral integrase (guided by the host cofactor LEDGF/p75) preferentially inserts the provirus into the body of actively transcribed genes rather than at random. This sim biases each landing site by a per-locus "transcriptional activity" value around the chromatin loop:
P(integrate at bin i) = activity(i) / Σ activity(j)
Four loci are marked as proto-oncogene TSS zones (analogous to LMO2, MECOM/EVI1 and similar genes implicated in real insertional-mutagenesis events). A landing site's genotoxic risk depends on the residual enhancer/promoter strength left in the vector's long terminal repeat (LTR) and its distance to the nearest oncogene TSS:
risk(i) = E · exp( -(d / 1.2)² )
E = enhancer strength (0–1), d = distance in loci to nearest oncogene
P(clonal dominance) = 1 − exp( −k · ΣE risk · (1 + 3·proliferation) ), k = 0.12
- SIN (self-inactivating) design — deletes the U3 enhancer/promoter from the 3′ LTR, so after reverse transcription both LTRs lose transactivating power. Toggling it on caps residual enhancer strength at 15% of the slider value, mirroring modern clinical lentiviral vectors.
- LTR-intact design — the historical (non-SIN, gammaretroviral-style) configuration used in the earliest X-SCID trials, where a strong LTR enhancer landing near LMO2 transactivated it and triggered leukemic clonal expansion in several patients.
- Proliferation pressure — models selective growth advantage in the transduced cell population; higher pressure lets even a modest genotoxic hit expand into a dominant, potentially malignant clone.
- Distance falloff — enhancer effects on a neighbouring promoter fall off sharply with genomic distance, so only integrations very close to a TSS carry meaningful risk.
This is a simplified, illustrative model of a real and well-documented safety mechanism in gene therapy — it is not a quantitative predictor for any specific clinical vector.