Retired Uricase Cassette Ranking (Computational, comp-022)¶
An active UOX cassette must produce correctly processed, localized, folded, and functional enzyme in the intended compartment. COMP-022 did not establish which A. oryzae configuration can do that.
Evidence boundary¶
COMP-022 is invalidated and non-runnable. Its CAI, RNA-structure, chaperone-load, promoter–signal-peptide, and ESM2 axes were not calibrated to one named biological outcome. No score, rank, shortlist, N-of-M tier, winner, component preference, gene-synthesis refinement, expression claim, processing claim, fold claim, secretion claim, activity claim, or safety inference survives.
The historical program enumerated a declared 6 promoters × 12 signal peptides × 10 codon variants × 60 scaffold labels = 43,200 rows. That is an inventory fact, not a validated parts universe or ranked build list.
Corrections to the historical record¶
- All four v1 top-cluster rows entered the retired v2 N-of-five ≥4 artifact tier; only one of the four entered N-of-five =5. The strict tier contained both PTS1-blocked and unblocked routes, so it did not confirm PTS1 masking.
- The historical file named
esmfold_pLDDT.csvcontains single-pass ESM2 log probabilities rescaled to 50–90. It is not ESMFold pLDDT or a fold-quality measurement. - The reviewed UniProt Q00511
sequence contains
NFSat residues 191–193, not the historicalNSSannotation, and terminates inSKL. N191Q would disrupt the N-X-S sequon, but COMP-022 established neither glycan occupancy nor a mutation preference.
These are corrections to an invalidated artifact, not rescued ranking results.
What remains worth testing¶
Direct secretion and GlaA-KEX2 processing remain distinct, unranked configurations. Promoter, signal peptide, codon design, terminal handling, propeptide, and glycosylation choices remain possible experimental factors; COMP-022 selects none of them.
Use matched exact constructs to measure transcript, processing and termini, localization, native or oligomeric state, fraction-specific intact active UOX, oxygen and peroxide behavior, host viability, and process retention. The construct-design page owns that matrix; validation §1.5 owns cassette characterization, and §1.33 owns the physiological UOX gate.
No successor ranking COMP is warranted until an exact cassette choice will drive near-term gene-synthesis spending.
Artifact: invalidated, non-runnable tombstone · computational experiment registry