Determine whether vulnerable codon-pair burden separates human transcripts by therapeutic desirability

Determine whether vulnerable-codon-pair burden and clustering distinguish human transcripts whose loss is therapeutically desirable from transcripts whose loss is toxic, thereby testing whether an exploitable selectivity separation exists.

Background

The proposed therapeutic strategy depends on collision-triggered decay preferentially affecting selected transcripts rather than broadly impairing translation. The paper identifies a structural concern: high translational flux is also characteristic of constitutively essential genes. It therefore leaves unresolved whether codon-pair statistics can separate oncogenic targets from essential transcripts in the human transcriptome; Milestone 1 specifies matched transcript sets and an AUC-based decision rule for testing this premise.

References

Whether an exploitable separation exists in the human transcriptome is an empirical question with a computable answer; it is the first milestone below.

— Kinetic Interference in Translational Control: A Path-Measure Framework for Collision-Triggered Transcript Decay  (2609.25536 - Segal, 22 Sep 2026) in Section 4.3, “The selectivity premise”