A, B and an O baby
At a village health camp, a newborn's blood group is recorded as O, the mother is A and the father is B. A relative insists there has been a mix-up at the hospital, arguing that 'an A mother and a B father can only produce AB children'. The medical officer explains that the ABO gene has three common alleles, I^A, I^B and i, and that both parents may be heterozygous; she writes the cross as I^A i × I^B i. The couple's first child, now four years old, is AB. Both children were tested twice with fresh antisera and gave the same result each time. The parents ask the probability that their next child will be group O, and whether the two children's groups are consistent with the same two parents. A neighbour also asks whether the husband could still be the father if his group were AB instead of B.
1. Use a Punnett square for I^A i × I^B i to judge the relative's claim.
Read the worked answer
Gametes: the mother can give I^A or i; the father can give I^B or i.
| I^B | i | |
|---|---|---|
| I^A | I^A I^B (AB) | I^A i (A) |
| i | I^B i (B) | ii (O) |
Result: AB : A : B : O = 1 : 1 : 1 : 1, i.e. 25% O.
Judging the claim: the relative is wrong on two counts.
- A group A parent can carry i, and a group B parent can also carry i. When both pass on i, the child is O (ii). Nothing is wrong with the hospital record.
- An AB child needs I^A from one parent and I^B from the other - which this cross also allows. So the same parents can produce both an AB child and an O child; the two results point in the same direction, not opposite ones.
The general lesson: you cannot read a person's genotype from the phenotype when the allele is dominant; you must consider the possible hidden allele.
2. Are the two children (one AB, one O) consistent with the same parents? Support your answer with a probability.
Read the worked answer
Yes, they are consistent.
Each fertilisation is an independent event with P(AB) = 1/4 and P(O) = 1/4 under the cross I^A i × I^B i.
For two children, either order works: AB then O, or O then AB. P = (1/4 × 1/4) + (1/4 × 1/4) = 2 × 1/16 = 1/8 (12.5%).
Interpretation: one chance in eight is uncommon but entirely ordinary; families of two children are far too small for 'unlikely' to mean 'impossible'. Consistency, however, is not proof of parentage - the cross only fails to exclude these parents. A genotype result can exclude a man, but blood groups alone are weak positive evidence, which is why courts and clinics do not treat a matching ABO group as identification.
3. If the husband's blood group were AB rather than B, would an O child still be possible? What would that mean for the neighbour's question?
Read the worked answer
No. A group AB father has the genotype I^A I^B and can pass on only I^A or I^B - never i. Crossing I^A I^B with the mother's I^A i gives:
- I^A I^A or I^A i → group A
- I^B i → group B
There is no ii combination, so a group O child is genetically impossible for that pair.
What this means: if the father really were AB, an O child would be strong grounds for ruling him out as the biological father, and the case would need proper DNA-based testing. Two cautions keep the conclusion honest:
- ABO results can be altered by recent transfusion or by a bone-marrow transplant, so samples and timing must be checked.
- Rare masking genotypes (for example the Bombay phenotype, where an hh individual types as O despite carrying A or B alleles) exist, so a single blood-group result is never treated as absolute proof by itself. The correct judgement is 'excluded, subject to confirmatory testing', not a flat accusation.