Source: https://guthlabs.ai/docs/instruments/punnett-based-trihybrid-multi-gene-probability-calculator/
instrument / verified-with-binding-errata
# Punnett-based Trihybrid/Multi-gene Probability Calculator
The current ledger marks this instrument verified with errata that must be applied at binding time.
Status: **VERIFIED_WITH_ERRATA**Lifecycle: **verified-with-binding-errata**
## Inspectable records
[01**Overview**Purpose and lifecycle]()[02**Mechanism**Method and assumptions]()[03**Implementation**Binding and validation]()[04**Qualification**Evidence and errata]()
## Ledger state
The current ledger marks this instrument verified with errata that must be applied at binding time.
**Area**: Biology, Medicine & Pharmacology (educational estimates only — every panel labeled not-medical-advice)
**Tranche**: tranche-5
**Evidence ID**: t5-116
**Ledger rows**: 1
No ledger note is attached.
## Review findings
The core product rule is standard under independent assortment for simultaneous per-locus events. The instrument cannot receive an unconditional PASS until it explicitly defines genotype versus phenotype inputs, decimal-versus-percent representation, validation of locus count, and the precise independence assumption. The educational and not-medical-advice labeling is appropriate, but it does not resolve those computational ambiguities.
### Formula and scope findings
- The product rule is correct only when the per-locus events are independent and the requested outcome is the simultaneous occurrence of one specified event at every locus.
- The specification alternates between genotype probability and phenotype probability. A phenotype probability can require summing multiple genotype probabilities before applying the product rule; the current formula does not define that case.
- The role of 'number of loci' is underspecified: it must equal the number of supplied per-locus probabilities, or the calculator must reject the input.
- The formula does not state whether probabilities are supplied as decimals from 0 to 1 or as percentages from 0 to 100.
### Unit assumptions
- Probabilities are dimensionless decimal fractions in the inclusive range 0 to 1; for example, 25% must be entered as 0.25.
- The number of loci is an integer of at least 3 and must equal the length of the per-locus probability list.
- Each per-locus probability refers to one specified genotype or one already-combined phenotype event.
- The loci assort independently; no linkage, epistasis, conditional dependence, or other dependence is included.
- The combined result is a decimal probability; multiply by 100 only when displaying it as a percentage.
## Boundary hardening
**BREAK_CLAIMED** · SELF_FLAGGED
The plain product rule has no division-by-zero trap and no genuine multi-regime formula branch, but an implementation without strict validation can return Infinity for extreme inputs and can accept values outside the probability domain. Percentage-versus-fraction ambiguity can also produce invalid results if units are not explicit.
## Errata and corrections
### Tranche errata
**combined_probability(_percent)** · vector 3
0 / 0
0.1 / 10
working literally writes 0.20×0.50=0
## Inspectable example
4 review vectors are recorded. One source-bound sample is shown below.
````
{
"inputs": {
"number_of_loci": 3,
"per_locus_probabilities": [
0.25,
0.5,
0.25
]
},
"expected_outputs": {
"combined_probability": 0.03125,
"combined_probability_percent": 3.125
},
"working": "P = 0.25 × 0.50 × 0.25\n0.25 × 0.50 = 0.125\n0.125 × 0.25 = 0.03125\n0.03125 × 100 = 3.125%"
}
````
## Related workflows
No source-defined multi-instrument workflow membership is recorded.
## Sources and evidence
These pointers bind the page to evidence. They do not grant authority or collapse verification into deployment.
- `VERIFIED_LEDGER.json · VERIFIED_WITH_ERRATA`
- `evidence-tranche5/tranche5-results.json`
- `BOUNDARY_HAZARD_CATALOG.json`