Why Pedigree Choice Matters
Family-based sequencing can help identify rare variants by showing how genetic differences are distributed among relatives with and without a phenotype. The family structure determines which comparisons are possible. A pedigree with several affected relatives may offer repeated opportunities to observe whether a candidate variant is shared with the phenotype. Parents and other relatives can help establish inheritance, clarify relationships among variant carriers, and provide context for interpreting a finding.
The biggest pedigree is not necessarily the most useful one. Some relatives may be unavailable, too young for a meaningful phenotype assessment, or connected through branches with little information about the phenotype. Conversely, a modest-sized family can be highly informative if several affected relatives and the relevant parents are available. Recruitment should therefore consider who is related to whom, who is affected, and who can actually be studied, rather than relying on a head count.
What Makes a Relative Informative?
The value of a relative depends on the question the study is trying to answer. An affected relative can show whether a candidate variant is shared among people with a similar phenotype. An unaffected relative can provide a contrast, especially if they are old enough that the phenotype would usually be apparent under the proposed disease model. A parent may help establish whether a variant was inherited and, in some cases, which parental copy was transmitted. A relative in another branch of the family may show whether a variant and phenotype co-occur across more than one set of descendants.
These roles are conditional on accurate data. Affected relatives should have a sufficiently similar and well-characterized phenotype. “Unaffected” should not be interpreted as meaningful evidence against a variant if the relative is too young, has limited clinical evaluation, or could develop the phenotype later. Likewise, apparent sharing can be misleading if relationships are recorded incorrectly or if genotypes are uncertain. Family size describes the pedigree; availability, phenotype quality, and relationship structure determine how much of it can contribute to the analysis.
For a single heterozygous variant in a parent, a simple Mendelian model provides a useful benchmark. Assume the other parent does not carry the variant, there is no de novo event or mosaicism, and each child’s transmission is independent. Each child then has probability \(1/2\) of inheriting the variant. This probability describes transmission, not the probability that a child will be affected. Whether a carrier develops a phenotype depends on the variant’s effects, penetrance, age, and other factors.
Worked Example: Counting Potential Transmissions in a Sibship
Suppose a hypothetical family has one heterozygous parent, one parent who does not carry a candidate variant, and five children who are available for sequencing. Three children are affected and two are unaffected. Under the Mendelian assumptions above, calculate the expected number of children who inherit the variant and the probability that at least one child inherits it.
For each child, the probability of inheriting the variant is \(1/2\). If \(X\) is the number of children who inherit it, then \(X\) has a binomial distribution with \(n=5\) and \(p=1/2\). Thus:
The expected value of 2.5 is an average over repeated families with this structure; it does not mean that half a child inherits the variant. To calculate the probability of at least one transmission, use the complement of no transmissions. Each child has probability \(1/2\) of not inheriting the variant, so:
The five available children offer a high chance of observing at least one transmission under this model. The three affected children may be particularly useful for assessing whether the variant is present among affected relatives, but their phenotype does not itself establish their genotype. Sequencing is needed to observe which children inherited the variant. The calculation also says nothing about whether carriers are expected to be affected.
Comparing Family Recruitment Strategies
Three broad structures often arise in family-based sequencing. A multiplex nuclear family has multiple affected relatives within a small family unit, often with parents available. It can provide several affected genotypes and direct information about parental inheritance. A larger extended pedigree may include affected relatives in multiple branches, allowing the variant’s distribution to be examined across more relationships. A small family with few affected relatives may offer less evidence of co-segregation, but sequencing available parents or older unaffected relatives can still clarify inheritance or provide useful contrasts.
These are not fixed rankings. An extended pedigree with many recorded members may have only one affected relative available for sequencing, whereas a smaller family might have several affected siblings and both parents available. Recruitment should prioritize the relatives who contribute to the actual analysis, not simply those who make a pedigree chart look large.
| Family feature | Potential contribution | Question to check |
|---|---|---|
| Several affected relatives | Allows comparison of variant sharing among people with the phenotype | Are their phenotypes sufficiently similar and reliably assessed? |
| Available parents | Can clarify inheritance and help check family relationships | Can both relevant parents be sequenced and interpreted? |
| Older unaffected relatives | Can provide a contrast to affected relatives | Are they old enough, and adequately assessed, to count as unaffected? |
| Relatives in separate branches | Can show whether a pattern is observed across distinct parts of the pedigree | Are the relationships and phenotype records clear? |
| Many listed but unavailable relatives | May add context to the pedigree, but not direct genotype evidence | Which relatives can actually contribute data? |
Worked Example: Choosing Between a Large and a Multiplex Pedigree
A hypothetical study team can recruit one of two families. Family A has 18 people listed in its pedigree, but only one affected person and one unaffected adult are available for sequencing; the parents are deceased. Family B has 8 people listed, including four affected siblings, both parents, and one older unaffected sibling, all available. Which family offers more direct opportunities to assess a candidate variant’s distribution?
Family B offers more direct opportunities. It has four affected relatives whose genotypes can be compared, both parents who can help clarify inheritance, and an older unaffected sibling who may provide a relevant contrast. Family A is larger on paper, but only two relatives can contribute genotype data, and neither parent is available. Its additional listed relatives do not become genetic observations simply by appearing in the pedigree.
This comparison does not prove that Family B will yield a causal finding, or that Family A has no value. Family B is better equipped for the specific task of comparing a candidate variant across affected relatives and assessing parental inheritance. The advantage depends on the stated availability and on the relatives’ phenotypes being accurately assessed. If the affected siblings have substantially different conditions, their presence would not automatically make the family more informative for one shared phenotype.
A Short Derivation: Why More Available Children Can Help
Under the simple transmission model, let \(m\) be the number of available children of a heterozygous parent. Each child independently has probability \(1/2\) of inheriting the variant. The probability that no child inherits it is the product of the probabilities of no transmission to each child:
The event “at least one child inherits the variant” is the complement of “no child inherits it.” Therefore:
For \(m=2\), this probability is \(1-(1/2)^2=0.75\); for \(m=5\), it is \(1-(1/2)^5=0.9688\). This derivation justifies one limited conclusion: more available children increase the chance of observing at least one transmission from a known heterozygous parent. It does not show that the chance of finding a disease-associated variant rises at the same rate, nor that the children provide independent phenotype evidence. Shared ancestry and shared environment remain relevant to interpretation.
Worked Example: Using Affected and Unaffected Relatives After a Candidate Is Found
In a hypothetical follow-up, a candidate variant has been observed in three affected siblings. It is also present in one of two available parents and absent from the other. An older unaffected sibling does not carry it; a younger unaffected sibling does carry it. Describe what this family pattern contributes without treating it as proof of causality.
The parent genotypes are consistent with the candidate being inherited from the parent who carries it. The three affected siblings sharing the variant makes the candidate’s distribution consistent with the phenotype in this family. The older unaffected sibling who does not carry it supplies a contrast that is also consistent with that pattern.
The younger unaffected carrier does not, by itself, contradict the pattern strongly: the child may not yet have reached an age when the phenotype would be expected. Even for an older unaffected carrier, interpretation would depend on the proposed penetrance and whether the phenotype had been carefully assessed. The family data support prioritizing the variant for further evaluation, but do not establish that it causes the phenotype. Incomplete penetrance, phenocopies, coincidental sharing, and other genetic or environmental factors may produce similar observations.
A Practical Recruitment Framework
Before enrolling relatives, make a compact recruitment plan. Record the phenotype status, age, relationship, and sequencing availability of each relative who could contribute. Then ask what each proposed sample adds: another affected comparison, a parental transmission, an older unaffected contrast, or information from another branch. A family can be valuable even when not every relative is sequenced, but the reasons for selecting particular relatives should be explicit.
Check whether affected relatives have a sufficiently similar, well-defined phenotype, and whether apparently unaffected relatives have been assessed appropriately for their age.
Identify which affected relatives, parents, older unaffected relatives, and relatives in other branches can actually be sequenced.
Prioritize samples that allow a meaningful affected-relative comparison, clarify inheritance, or provide a credible phenotype contrast.
Assess whether a candidate’s observed distribution is compatible with the phenotype, while allowing for age, penetrance, phenotype misclassification, and other explanations.
Common Pitfalls and Why the Choice Matters
A common mistake is to reason, “This pedigree has 20 members, so it is more informative than a family with six.” That conclusion confuses the number of relatives recorded with the number of usable genetic and phenotype observations. The correction is to count the affected relatives, relevant parents, and appropriate comparison relatives who are available, and to ask what each can contribute.
Another mistake is to treat every unaffected relative as evidence against a candidate. An unaffected child may be too young for the phenotype to appear; a relative may have incomplete clinical evaluation; or the disease mechanism may allow carriers to remain unaffected. Conversely, finding a variant in several affected relatives is not proof of causality, because relatives share many variants and family-specific factors.
Family-based sequencing matters because a thoughtful choice of relatives can make inheritance and phenotype patterns easier to evaluate. The most useful pedigree is not defined by size alone. It is the pedigree in which the available relatives provide relevant, reliable comparisons for the rare-variant question being asked.
Check Your Understanding
Use the assumptions and distinctions in this tutorial to answer the following questions.
- A heterozygous parent has four available children, and the other parent is a noncarrier. Under independent Mendelian transmission, what is the probability that at least one child inherits the variant?
- Why might a smaller family with several affected siblings and both parents available be more useful than a larger pedigree with only one affected relative available?
- What information can an available parent contribute when evaluating a candidate variant in affected relatives?
- Why may a young unaffected relative provide weak evidence against a candidate variant?
- A candidate variant is present in all three affected relatives studied. Why does that observation not establish that the variant causes the phenotype?