Page actions
instrument mechanism / verified-clean
Primer Melting Temperature (Tm) Calculator mechanism
The inspectable method record for Primer Melting Temperature (Tm) Calculator, kept separate from implementation and qualification.
Model scope
The verdict is NEEDS_REVIEW because two alternative algorithms are named but the nearest-neighbor implementation is not fully defined. Vectors 1-3 exercise exact standard Wallace-rule coefficient entries, including zero-GC and high-GC boundary behavior. The nearest-neighbor vector uses the stated SantaLucia 1998 convention and explicitly supplies every table value needed for reproduction.
- The Wallace rule Tm=4(G+C)+2(A+T) is a standard rough estimate for short, relatively simple oligonucleotides, but its applicability limits should be stated explicitly; it is not generally reliable for long primers, unusual salt conditions, mismatches, secondary structure, or modified bases.
- The salt-adjusted nearest-neighbor method is underspecified: it must define the exact SantaLucia 1998 parameter convention, salt model, oligo concentration convention, self-complementarity treatment, terminal AT corrections, initiation terms, and whether divalent ions are supported.
- The nearest-neighbor formula must specify whether concentration is total strand concentration and whether the non-self-complementary correction ln(Ct/4) or a self-complementary correction is used.
- GC% is not mathematically defined in the specification; the usual definition is 100*(G+C)/sequence length, with ambiguity for empty or ambiguous-base sequences.
- The specification calls the instrument a lookup-table engine but does not state how missing or ambiguous nucleotide symbols are handled.
Units and conventions
- Sequence contains only uppercase or case-normalized A, C, G, and T bases; ambiguous symbols, gaps, and modified bases are rejected unless separately modeled.
- Wallace-rule coefficients are 4 degrees C per G/C base and 2 degrees C per A/T base; the result is in degrees C.
- GC% is reported as a percentage, not a fraction.
- For the nearest-neighbor vector, SantaLucia-style enthalpies are in kcal/mol, entropies are in cal/(mol*K), R=1.9872 cal/(mol*K), oligo concentration is total strand concentration in mol/L, and monovalent sodium concentration is in mol/L.
- The nearest-neighbor salt correction used here is +16.6*log10([Na+]) degrees C, with [Na+] in mol/L; this is a monovalent-salt approximation and does not independently model Mg2+ or dNTP binding.
- The nearest-neighbor example assumes a non-self-complementary duplex and uses Tm(K)=1000*DeltaH/(DeltaS+R*ln(Ct/4)), followed by conversion to degrees C and the sodium correction.
Inspectable vector
One of 4 source-bound review vectors:
{
"inputs": {
"sequence": "ATGCGTACGATCGATCGTAC",
"method": "Wallace rule"
},
"expected_outputs": {
"Tm_C": 60,
"GC_percent": 50
},
"reference_data": {
"wallace_GC_coefficient_C_per_base": 4,
"wallace_AT_coefficient_C_per_base": 2,
"source_convention": "widely tabulated standard Wallace-rule values"
},
"working": "Length = 20. Count G+C = 10 and A+T = 10. Tm = 4*10 + 2*10 = 40 + 20 = 60 deg C. GC% = 100*(10/20) = 50%."
}Failure boundary
BREAK_CLAIMED · SELF_FLAGGED
The specification permits hazardous ambiguity: an empty sequence can produce a misleading raw Wallace value of 0 °C while GC% is undefined, and the nearest-neighbor branch cannot produce a numeric result without the required parameter table and explicit method, unit, and convention validation.
Other records
Sources and evidence
These pointers bind the page to evidence. They do not grant authority or collapse verification into deployment.
VERIFIED_LEDGER.jsonevidence-tranche7/tranche7-results.jsonBOUNDARY_HAZARD_CATALOG.json