eduKateSG · Genetics · The Gold Standard of Education
Map a mutation with a deletion, then prove what the result can and cannot mean
Choose the explanation you need, or follow all twelve chapters in order.
Full chapter index · Independent practice · Genetics subject guide
Deficiency mapping looks simple on a diagram: cross a mutation to strains carrying known deletions, find the deletion that fails to complement, and place the mutation inside the missing interval. The real experiment is more careful. A deletion changes dosage for every locus it removes. The scored offspring may be sick, inviable or hard to distinguish. A failure to recover the expected class may therefore reflect allelism, dosage sensitivity, a second mutation, a balancer, a scoring error or a cross that never produced the intended genotype.
This guide turns that compressed diagram into a defensible chain of reasoning. You will define the genotype that provides the test, keep chromosome structure separate from phenotype, read overlapping deficiencies as interval evidence, and use rescue or molecular data to close the gap between mapping and identification. The worked examples use fictional loci and counts so that the reasoning is visible without pretending to report a real organism or experiment.
Deficiency mapping is especially useful when recombination is suppressed or when a rapid regional test is needed. It is not a universal replacement for recombination mapping, sequencing or direct functional validation. The aim is a bounded conclusion: what interval is supported, what alternatives remain, and which next cross or assay has the highest information value.
Continue from The Gold Standard of Genetics within The Gold Standard of Education: learning that transfers into life and the How X Works learning library.
Full chapter index
See the mechanism · Chapters 1–2
Build the worked model · Chapters 3–5
Test what changes · Chapters 6–8
Check evidence and limits · Chapters 9–10
Practise and continue · Chapters 11–12
A deficiency is a chromosome with a defined segment absent. In a diploid, an individual carrying one normal homolog and one deficient homolog has only one copy of every locus inside that interval. This is hemizygosity, not ordinary heterozygosity. The distinction matters because the test mutation has no wild-type partner at a deleted locus. If the mutation lies inside the deleted interval and is recessive, the mutant phenotype may be exposed in the mutation/deficiency offspring.
Outside the deleted interval, the deficiency chromosome still carries alleles. At a locus beyond either breakpoint, a wild-type allele on the deficiency chromosome can complement a recessive test mutation. The map is therefore read through breakpoints: a failure with one deficiency and complementation with an overlapping neighbour can locate the mutation between their endpoints. The resolution is set by the pattern and precision of those endpoints, not by how confidently the phenotype was scored.
A deficiency also changes more than the target locus. It removes neighbouring genes and regulatory elements, may alter chromatin near a breakpoint, and can reduce viability through combined dosage effects. Some deletion heterozygotes have a visible phenotype even without the test mutation. Begin by recording the deficiency-alone phenotype and viability in the same background and conditions. Otherwise a deletion effect can be mistaken for failure to complement.
The first transfer check is verbal. Replace “the deficiency contains the mutation” with the more exact statement: “the mutation maps within the interval absent from this deficiency, subject to the controls.” A deletion is missing DNA; it does not contain the mutant allele. Precise language prevents a precise map from being built on a conceptual reversal.
Evidence check: For What a deficiency changes, preserve the observation before the interpretation. Record the informative genotype and every sibling class. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
A standard complementation test asks whether two recessive mutations can supply one another’s missing function in a trans-heterozygote. Deficiency mapping asks a related but asymmetric question. One chromosome carries the test mutation; the other lacks an interval. If the target gene lies in that interval, the offspring has no functional copy from the deficiency chromosome. A mutant phenotype is then called failure to complement the deficiency.
Write the informative genotype before looking at a result. Suppose m is a recessive mutation and Df(A–D) deletes loci A through D. The critical class is m/Df, not every F1 animal. If m is at locus C, the genotype has mutant C on one homolog and no C on the other. If m is at locus F, outside the deletion, the deficiency chromosome may still carry C+ and F+; at F the offspring is m/F+ and should usually look wild type.
“Usually” is deliberate. Non-complementation can arise between mutations in different genes when their products interact, when total pathway dosage crosses a threshold, or when the background carries another lesion. Intragenic complementation can also make two alleles of one gene appear to complement. The existing guide to complementation tests and intragenic exceptions explains those cases. Deficiency mapping adds the complication that many genes are simultaneously reduced to one copy.
Use three labels in the notebook: observed phenotype, inferred complementation state, and map implication. Keeping them separate makes it easier to revise the inference when penetrance, viability or genotype confirmation changes. The phenotype is data. “Fails to complement” is an interpretation under a model. “Inside the deleted interval” is a further map conclusion that requires the deficiency and test genotypes to be correct.
Evidence check: For Complementation under hemizygosity, preserve the observation before the interpretation. Record the deficiency-alone phenotype in the matched background. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
CHAPTER 3 OF 12 · Build the worked model
3. Design the cross around an informative class
The best cross is designed from the offspring backward. Identify the genotype that can answer the question, the markers that distinguish it, and the sibling classes that test whether the cross behaved. In many systems a deficiency is maintained over a balancer or marked homolog because the deletion is homozygous lethal. The test mutation may also be balanced. The cross can therefore produce several visually similar F1 classes.
Make a cross table before mating. List the parental homologs, gametes, expected F1 genotypes, visible or molecular markers, predicted phenotype, and whether each class is viable. If the informative m/Df class cannot be distinguished from m/balancer or wild-type/Df, the experiment needs another marker or genotyping step. Do not plan to infer a genotype from the phenotype being tested; that creates circular evidence.
Positive and negative controls answer different questions. A wild-type chromosome crossed to the deficiency tests the deletion-alone background. A known mutation inside the deletion tests whether the deficiency can reveal a recessive allele under the chosen conditions. A known outside mutation tests complementation beyond the breakpoint. Reciprocal crosses may expose maternal, paternal or cytoplasmic effects, although reciprocal equality does not prove their absence in every context.
Scale the cross for the rarest informative class, not for the total number of offspring. If half the progeny are excluded by a balancer and another fraction is inviable, one hundred scored offspring may yield only a handful of m/Df individuals. Record embryos, larvae or seeds lost before the usual scoring stage when possible. Missing informative progeny can itself be evidence, but only when the expected class and recovery window are explicit.
Evidence check: For Design the cross around an informative class, preserve the observation before the interpretation. Record the cross table and planned genotype marker. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
Consider a fictional chromosome region ordered L1—L2—L3—L4—L5—L6. Four deficiencies have molecularly checked intervals: Df-A removes L1 to L3, Df-B removes L2 to L4, Df-C removes L3 to L5, and Df-D removes L5 to L6. A recessive mutation amber causes short bristles when no functional copy of its gene is present.
Cross amber to each deficiency and genotype the informative F1 class. The invented results are: amber/Df-A is wild type in 96 of 100 confirmed animals; amber/Df-B shows short bristles in 91 of 98; amber/Df-C shows short bristles in 94 of 101; and amber/Df-D is wild type in 103 of 105. Small departures reflect ordinary scoring noise in this teaching example.
The intersection of the non-complementing intervals is L3–L4. The complementing Df-A removes L1–L3, while Df-D removes L5–L6. If endpoints and locus order are exact, the combined pattern narrows amber to the portion shared by B and C but not convincingly exposed by A or D. Notice that the answer is an interval, not “gene L4.” Breakpoint uncertainty and the partial overlap at L3 must be carried forward.
Now change one result: amber/Df-C is not recovered at all, while sibling classes appear. Do not automatically score that as non-complementation. It could mean the combination is lethal, the genotype marker failed, or the cross produced too few informative offspring. Repeat at an earlier developmental stage and genotype non-survivors if feasible. “No animals” and “animals with the mutant phenotype” are biologically different observations.
The table attached to this chapter separates interval geometry from phenotype. A map becomes persuasive when another reader can reproduce the intersection and see which observations determine each boundary.
| Deficiency | Deleted interval | Confirmed informative F1 | Observed result |
|---|---|---|---|
| Df-A | L1–L3 | 100 | Mostly wild type |
| Df-B | L2–L4 | 98 | 91 short-bristle |
| Df-C | L3–L5 | 101 | 94 short-bristle |
| Df-D | L5–L6 | 105 | Mostly wild type |
Evidence check: For Worked map with overlapping deficiencies, preserve the observation before the interpretation. Record the exact overlap and every breakpoint uncertainty. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
Recessive mapping assumes that one normal copy is enough. Haploinsufficient loci violate that assumption: a wild-type/deficiency individual may show a phenotype because one copy does not supply sufficient product. If the deficiency alone produces short bristles, then short bristles in amber/Df do not identify amber’s locus. The result needs comparison with matched wild-type/Df siblings, ideally under the same age, temperature, nutrition and scoring criteria.
Combined dosage effects can be subtler. The deficiency may reduce a modifier or pathway partner so that a weak test allele becomes visible even when its own locus lies outside the deletion. This is non-allelic non-complementation. It can reveal useful biology, but it is not the simple positional inference. A duplication spanning the candidate interval, a small rescue construct or an independent deficiency with different neighbouring genes can help distinguish target exposure from regional dosage interaction.
Dosage can also change penetrance rather than create an all-or-none phenotype. Compare distributions, not only labels. For a quantitative trait, predefine a measurement and plot confirmed genotypes. If wild-type/Df shifts slightly, amber/+ shifts slightly, and amber/Df shifts strongly, the pattern may indicate interaction. Calling only the last class “mutant” discards the evidence needed to understand why.
The practical limit is important: a clean complementation result is often more secure than a failure. When a deletion-bearing chromosome supplies a working copy of the outside locus and the offspring is healthy, the mutation is less likely to lie within the deleted interval. Failure can have more alternative causes. Treat asymmetry as part of the method rather than as an inconvenience.
Evidence check: For Dosage and haploinsufficiency, preserve the observation before the interpretation. Record the baseline effect of reduced dosage. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
CHAPTER 6 OF 12 · Test what changes
6. Breakpoints are measurements, not magic borders
Older deficiencies were often bounded by visible markers or cytological bands. Modern resources may supply sequence coordinates, but coordinates still depend on a reference assembly and breakpoint assay. A breakpoint can interrupt a gene, remove a distant enhancer, create a fusion or be accompanied by an inversion or insertion. “Deletes 2.1–2.8 Mb” is a measured model of a rearranged chromosome, not a perfect rectangle cut from the reference.
Record the organism, strain, deficiency identifier, reference assembly and source of breakpoint information. Two resources can use different assemblies or update an interval after sequencing. Lifted coordinates may shift near repeats. If a candidate gene lies close to a boundary, inspect the primary characterization rather than relying on a browser thumbnail.
Overlaps should be computed in the same coordinate system. If Df-B ends at 2.64 Mb and Df-C starts at 2.58 Mb, their shared interval is about 60 kb only if both coordinates refer to comparable breakpoint definitions. Uncertainty may be larger than the apparent overlap. Use language such as “consistent with the shared region” until a smaller rearrangement or molecular test resolves it.
A changed-context check makes this concrete. Suppose a new assembly places one candidate gene across the Df-B breakpoint. The phenotypic data have not changed, but the candidate list has. Re-run the interval intersection and document the assembly version. Good mapping is reproducible reasoning attached to explicit map data, not a conclusion detached from its coordinate history.
Evidence check: For Breakpoints are measurements, not magic borders, preserve the observation before the interpretation. Record the assembly version and breakpoint assay. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
Deficiency mapping observes the offspring that survive to the scoring point. If the informative genotype dies earlier, the visible sample is selected. The attached fictional recovery table shows how a cross that begins with equal zygotic expectations can end with a severe shortage of m/Df adults. Counting only adults would make the class appear absent without locating when or why it was lost.
Use sibling ratios as an alarm, not as proof. A deficit of the informative class can signal lethality, reduced fertility, marker misclassification, segregation distortion or parental genotype error. Stage-specific counts, embryo hatching, larval survival and direct genotyping can separate these possibilities. The relevant denominator is the number of offspring in which the genotype could have been detected.
Temperature, diet and microbial exposure may alter survival or penetrance. If a failure appears only at a restrictive temperature, report conditional non-complementation rather than a universal map fact. Conditionality can be biologically valuable; it may reveal a sensitised pathway. It also means that another laboratory must reproduce the environment as well as the cross.
Do not rescue a disappointing ratio by merging categories after seeing the data. Define viable, affected, ambiguous and unscorable classes in advance. Preserve the raw counts and images or instrument output where appropriate. A map interval derived from transparent exclusions is stronger than a cleaner result created through undocumented filtering.
| Class | Expected zygotes | Recovered early | Scored adult |
|---|---|---|---|
| m/Df | 120 | 88 | 31 |
| m/Bal | 120 | 112 | 103 |
| +/Df | 120 | 109 | 96 |
| +/Bal | 120 | 116 | 108 |
Evidence check: For Recovery bias and lethal combinations, preserve the observation before the interpretation. Record the stage at which missing offspring disappear. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
CHAPTER 8 OF 12 · Test what changes
8. When non-complementation does not mean allelism
The central misconception is “failure to complement proves the same gene.” Even ordinary two-allele tests have exceptions. A deficiency intensifies them because many loci are hemizygous. A mutation outside the interval may interact with reduced dosage inside it; a dominant or semidominant allele may show regardless of location; a second-site mutation carried with the test allele may be the true interacting lesion.
Background control is therefore part of mapping. Outcross the test mutation, use independently derived alleles, and compare more than one deficiency spanning the same candidate region when possible. If two molecularly different deletions with a common small overlap fail while flanking deletions complement, the positional argument strengthens. If only one stock fails, suspect a stock-specific modifier or inaccurate breakpoint before declaring a new locus.
Transvection, regulatory interactions and multimeric proteins can complicate complementation in some organisms. The correct response is not to memorise every exception before doing a cross. It is to build orthogonal evidence: phenotype, genotype, interval, independent allele, rescue and sequence. Each line rules out different alternatives.
Use the phrase “maps by deficiency to” for the interval-level result and reserve “is an allele of” for evidence that identifies the gene. This linguistic discipline keeps the conclusion proportional. It also makes a surprising rescue or sequence result easier to integrate; the earlier statement remains accurate even if the causal gene was not the first candidate.
Evidence check: For When non-complementation does not mean allelism, preserve the observation before the interpretation. Record stock-specific modifiers and independent alleles. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
After an interval is defined, list every coding gene, non-coding feature and relevant regulatory element within it. Prioritisation can use expression, known pathway roles and phenotype similarity, but those are filters, not proof. A famous candidate can attract attention while a poorly annotated gene is ignored. Keep the full interval list available so that negative tests do not vanish.
Sequencing a candidate allele can reveal a plausible variant, yet plausibility still needs comparison with parental strain and population background. A missense change in a conserved residue is stronger than a common synonymous variant, but neither automatically establishes causality. Independent alleles with related phenotypes, precise genome editing, or rescue with a wild-type copy can connect variant and phenotype.
Rescue must be interpreted carefully. A large duplication may restore several deleted genes, and an overexpressed transgene can bypass a pathway rather than replace the causal function. Map the rescue fragment, control its copy number or expression where possible, and show that the control construct does not rescue. Reciprocal evidence—recreating the mutation in wild type and repairing it in the mutant—is especially strong.
The decision sequence is efficient when each step narrows uncertainty: overlapping deficiencies define a region; smaller rearrangements or recombinants refine it; sequencing nominates variants; rescue or editing tests causality. Jumping directly from a broad deletion to one attractive gene skips the evidence that protects the conclusion from confirmation bias.
Evidence check: For From interval to gene, preserve the observation before the interpretation. Record the complete candidate list and rescue boundaries. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
Before calling the map, ask five questions. Was the informative genotype directly identified? Did the deficiency-alone control establish its baseline phenotype? Were the breakpoints checked in a named coordinate system? Did at least one independent interval or allele support the boundary? Were survival and ambiguity reported rather than silently excluded?
Set a stopping rule before expanding the cross. For a binary phenotype, define the minimum number of confirmed informative offspring and the tolerated ambiguous fraction. For variable penetrance, define the effect or interval precision that would justify the next assay. More offspring reduce sampling noise but do not repair a confounded genotype or an unknown breakpoint.
Distinguish a map failure from an uninformative result. If every available deficiency produces its own severe phenotype, the method may not be suitable for that trait. If no deficiency covers part of the region, a negative screen cannot exclude that gap. If all tested deletions complement, the mutation may lie outside the panel, be rescued by a duplication in a stock, or require a condition not used.
An auditable result includes the cross scheme, stock identifiers, genotype assay, raw class counts, scoring window, exclusion rules, interval calculation and unresolved alternatives. These details are not administrative weight. They are what allow a reader to distinguish a true positional signal from a convenient story.
Evidence check: For Evidence checks and stopping rules, preserve the observation before the interpretation. Record raw counts, exclusions and the stopping rule. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
Practice case one: Df-1 removes genes A–D, Df-2 removes C–F, Df-3 removes E–H. A recessive mutation fails with Df-1 and Df-2 but complements Df-3. The simplest interval is C–D. Now add the fact that wild-type/Df-1 shows the same phenotype at 40 per cent penetrance. The Df-1 result becomes less informative; Df-2 carries more weight, and a second overlapping deficiency or direct genotype-phenotype comparison is needed.
Practice case two: a test mutation complements two deletions but m/Df offspring are only one tenth as common as expected. The correct first conclusion is not that the mutation lies outside both deletions. Determine whether the informative genotype is lost before scoring. Genotype embryos or an earlier stage and inspect sibling ratios. Survival bias can turn a biological interaction into an apparent clean complement.
Practice case three: sequence coordinates place candidate X barely outside a deletion, but a rescue fragment containing X restores the phenotype. Recheck the breakpoint and rescue boundaries. The deletion may remove a regulator of X, the assembly mapping may be wrong, or overexpression of X may bypass the defect. The tension between map and rescue is a reason to design another test, not a reason to discard whichever result is less convenient.
Transfer the method beyond genetics. A deficiency panel is a set of structured perturbations. The general skill is to infer location from overlapping changes while tracking side effects and observation filters. The same reasoning appears in fault isolation, ablation experiments and missing-data diagnosis: overlap can localise a cause only when each perturbation is accurately characterised and alternative pathways are controlled.
Evidence check: For Independent practice and transfer, preserve the observation before the interpretation. Record which observation would reverse the interval. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.
Does one failing deficiency identify the gene?
No. It usually identifies a candidate interval and may also expose dosage interactions or stock problems. Use overlapping deficiencies, accurate breakpoints and an independent gene-level test.
Is no recovered m/Df class the same as a mutant phenotype?
No. Absence may reflect lethality, failed production or failed recognition. Locate the missing class with earlier counts and direct genotyping.
Why can a deficiency outside the target gene change the phenotype?
It reduces dosage for many other genes. A pathway partner or modifier can interact with the test allele, producing non-allelic non-complementation.
What is the most useful next experiment after a clean overlap?
Choose the test that removes the largest live alternative. That may be a smaller deficiency, a duplication rescue, sequencing of all interval candidates, or an independent allele. The best next step depends on whether interval width, stock identity or causal-gene proof is the current bottleneck.
Authoritative reading includes the NCBI Bookshelf chapter Mapping with deficiencies and duplications, the WormBook review Complementation, and the chapter Karyotype, ploidy and gene dosage. These sources explain the method and its organism-specific complications; stock-centre records and the primary breakpoint paper remain necessary for any real deficiency.
Return to The Gold Standard of Genetics for the wider subject map. Use chromosomal inversions and recovered recombination when rearrangements change the offspring that appear, and gene conversion and non-Mendelian tetrad ratios when segregation itself departs from the simplest model. The next useful move is always the same: state the genotype, state the interval, state the alternative explanations, then choose the smallest experiment that separates them.
Evidence check: For FAQ, sources and the next useful move, preserve the observation before the interpretation. Record the claim that is supported and the alternatives still live. Then write the inference, one credible alternative and the additional observation that would separate them. This short audit keeps fluent terminology from outrunning the evidence actually available.