Inspect a real fruit fly reconstruction, choose inherited copies, exchange chromosome segments and test why a probability does not promise an exact family. Follow linked markers from parent to gamete to offspring.
Enable JavaScript to change the conditions and run the interactive experiment.
Offspring inherit combinations of DNA. The arrangement of variants on chromosomes helps determine which combinations can travel together.
The parent keeps the same four marker copies, but their chromosome arrangement changes. Must the offspring-pair probabilities stay the same?
Phase changes the pairing of variants while preserving individual copy totals.
The source-derived fruit fly gives the lesson an anatomical anchor: a body, wings, eyes and six legs. It comes from a prepared female specimen used in NeuroMechFly research. Our controls do not reshape that specimen into supposedly measured mutant offspring. The genetic workbench is a separate, declared model.
Start with two known +/vg parents. Each supplies one allele at this marker through a gamete. Combining their contributions gives +/+, +/vg or vg/vg. Under the selected recessive mapping, only vg/vg produces the vestigial-wing class. A parent’s reference-wing appearance does not by itself reveal both copies.
The four equally likely parental-copy combinations give genotype probabilities 1/4, 1/2 and 1/4. Grouping the specified wing classes gives 3/4 and 1/4. Four independent offspring need not include one of every possibility. The sample activity retains every generated outcome, including surprising batches.
The body marker and wing marker lie on the same chromosome. A parent can carry the variants together on one homolog, or one variant on each homolog. The arrangements contain identical per-marker totals but different pairs. This arrangement is phase: together/cis or split/trans.
After DNA replication there are four chromatids in the paired homologs. The demonstration selects one chromatid from each homolog and exchanges a segment. Two marker pairs change while the total copies are conserved. The diagram then follows one selected gamete product, not four equally functional eggs.
In the testcross, the other parent always supplies (b, vg). This makes the sampled marker pair readable from the offspring’s tracked class under our specified recessive mapping. Compare exact probabilities with a saved sample, then trace any offspring back to its two gametes.
An allele is a version at a genetic locus. A genotype describes the inherited copies being tracked; a phenotype is an observed characteristic. Neither two markers nor a single outward feature specifies a whole organism. Dominance is a relationship between selected alleles and a measured trait in a context, not a measure of strength, quality or frequency.
Morgan’s 1916 example assigns 17% total recombinant combinations to this female-fly cross. For a together-phase parent, the four canonical pairs (+,+), (+,vg), (b,+), (b,vg) receive probabilities 83/200, 17/200, 17/200 and 83/200. Split phase reverses which pairs are common. Each individual variant still has probability 1/2.
One selected nonsister exchange makes two recombinant chromatids among four. That count does not establish that 17% of cells undergo the illustrated event. The historical fraction is a separate ensemble description. Our model does not derive a crossover frequency from marker spacing or draw multiple exchanges, gene conversion or every meiotic product.
Standard Drosophila melanogaster male meiosis lacks the usual meiotic crossing over. Our male comparison therefore supplies only the two parental marker combinations, with probability 1/2 each. This is a species-specific comparison, not a rule about human fathers or all male animals. The tester remains fixed when the modeled parent changes.
The seeded testcross uses one uniform pseudorandom draw per offspring and exact integer probability weights summing to 200. A generated run stores its phase, mode, seed and sample size. Later control changes prepare a new run without reinterpreting the saved results. The expected count N × p can be fractional; observed simulated counts are integers.
Morgan and Lynch’s 1912 recent-cross F2 count lists 1,081 reference-wing and 280 vestigial-wing adults; the reciprocal-cross F2 lists 1,213 and 273. These are preserved source observations, not generated data. An adult sample can be affected by viability and observation as well as the inherited combinations. The original investigators considered viability.
FlyBase now names the historical body-color b gene Adc, on chromosome arm 2L. The vg gene lies on 2R and contributes, with other factors, to wing and haltere identity. The schematic does not convert arm-local sequence coordinates into a distance or apply the same phenotype mapping to every possible allele.
Voigt and colleagues measured normalized vg mRNA in pooled adult samples at 17 °C and 28 °C. Their reported adult-male ratios are 1.76 for Sweden and 2.28 for France. Those are expression ratios, not wing-size multipliers or individual forecasts. The source has an unresolved ZI/ZK population-label mismatch; the affected ratios are not assigned here.
Recombination and segregation help explain new combinations. Many human traits involve numerous genes, development and environments. This fly example introduces the mechanism while deliberately stopping short of predicting anyone’s appearance, ancestry or health.
A specified tester can make otherwise hidden allele combinations easier to trace. The testcross workbench shows why knowing the parental genotypes and phase matters before interpreting offspring counts.
An anatomical reconstruction, a counted adult sample, a normalized expression measurement and an ideal probability table answer different questions. Preserve their units, assumptions and provenance when comparing them.
Put a + and a vg card in each parent’s envelope. Draw one from each without looking, record both, then replace and mix. Repeat twelve times; tally genotypes before grouping wing classes.
Make two (+,+) and two (b,vg) chromosome cards. Draw a whole card from one envelope, pair it with a fixed (b,vg) tester, record, replace and mix. This no-exchange paper model produces only original pairs.
Lay out two duplicated homologs. Cover the right-hand marker on one card from each homolog with exchanged labels. Identify the two changed pairs. This shows a selected exchange, not a 17% frequency.
Start with two (+,vg) and two (b,+) cards. Repeat no-exchange draws. Compare the possible pairs without changing the individual marker totals.
Write what was possible, what each model predicted, and what your paper trials actually produced. Explain which assumption changes if you draw the two markers independently.
How can the same individual marker totals lead to different combinations being passed on?
An original paper model, not chromosome chemistry or a human trait predictor. Use large cards and tear paper if needed. No family histories, biological samples or insect breeding.
Yes, each can supply a vg copy. The offspring inherits one copy from each parent.
The probabilities do not guarantee an exact count. Independent draws can produce many different finite counts.
Which two markers share each homolog. Phase describes marker arrangement across the homologs.
It would discard the assumption that the markers travel together. A whole chromosome card preserves the linked pair.
No, a selected event and an ensemble rate are different. The chosen demonstration does not derive the historical rate.
Only the two parental marker pairs are produced in this model. This is a species-specific meiosis comparison.
Did survival or observation differ before adults were counted? Adult counts can include processes after fertilization.
Expression differed in that measured context; this does not predict individual appearance. The measured quantity is mRNA abundance, not anatomy.
National Human Genome Research Institute glossary. The fly workbench tracks a selected pair, not every chromosome or meiotic product.
NHGRI · meiosisPrimary institutional genetics explanation; the segment animation is explicitly schematic.
NHGRI · crossing overSupports the boundary against using a two-marker fly cross to predict human appearance.
NHGRI · polygenic traitCurated Drosophila record. The historical symbol b is retained on teaching cards to match the historical cross; modern name Adc is taught alongside it.
FlyBase · Adc, formerly black/bCurated record. Does not imply that all vg mutations are equivalent, recessive or confined to the displayed phenotype.
FlyBase · vestigial/vgA Critique of the Theory of Evolution, Figure 66 and accompanying text. Public domain in the US. Historical quantitative example retained; obsolete historical molecular explanations are not reused as modern fact.
T. H. Morgan (1916) · original linkage exampleThe Linkage of Two Factors in Drosophila That Are Not Sex-Linked, printed page 177. Counts are transcribed as source observations and kept separate from generated simulations.
Morgan & Lynch (1912) · counted fliesGenes 10(7):498, Figure 1 and Table 3. CC BY 4.0 figure retained in full. Three biological replicates; technical repetitions are not independent biological samples. ZI/ZK label conflict disclosed.
Voigt, Erpf & Stephan (2019) · vg expressionNature Methods, source Methods. The paper has separate copyright; no restricted paper figure is copied. The geometry is reused under its repository license.
Lobato-Rios et al. (2022) · NeuroMechFlyPinned commit; 65 mesh parts and 334,525 triangles retained. Public provenance records hashes, scale and coordinate conversion. Authored colors and inspection pose are identified separately.
NeuroMechFly · source geometry and licenseIndependent subject review is pending.
Read the sources and model assumptions
Interactive explorationTake apart a human ribosome, follow a message three letters at a time, then change a DNA base and discover what the peptide does.
Release a few virtual balls, then a thousand. Change the chance of a right turn and compare the results with a mathematical prediction.
Interactive explorationMatch two kinds of fiber to the microbes that can use them. Look through a bacterial surface, inspect a real enzyme holding a sugar chain, and sort the different kinds of products that microbes can make.