1. A locus is a particular position on a chromosome. For diploid individuals, there are two versions of every locus, one inherited from the mother, the other from the father. An allele is one of the potentially very many DNA sequences that may appear at that position on the chromosome. Each allele may have a different phenotypic effect. A gene may refer to either a locus (the eye colour gene), or an allele (the blue eye gene), and ironically enough is best avoided in discussions of genetics because of this confusion.
  2. A hexaploid species such as wheat may have one of seven genotypes at each locus: AAAAAA, AAAAAa, AAAAaa, AAAaa, AAaaaa, Aaaaaa or aaaaaa.
  3. There will be no YY individuals in the population, since they will all die. Fertile matings can only occur between XX and XY individuals:

     

    X

    X

    X

    XX

    XX

    Y

    XY

    XY
    The sex ratio as seen from this Punnet square is obviously 1:1.
  4. F2 = F1 RrYy x test-cross-parent rryy = 1 RrYy : 1 rrYy : 1 Rryy : 1 rryy. A 1:1:1:1 ratio, which is th whole reason for a test cross, since any deviations from 1:1:1:1 are very easy to spot. Note that the test cross parent need not be one of the parental strains. In this case it is not because the R and Y alleles were in 'repulsion' phase: the parental strains were RRyy and rrYY, not in 'coupling' phase (RRYY x rryy).
  5. If the genes were closely linked, there would be an excess of the parental arrangements of the genes: R with y (Rryy) and r with Y (rrYy). There would be fewer of the recombinant arrangements, R with Y (RrYy) and r with y (rryy).
  6. Assuming no linkage, we would expect there to be a 1:1:1:1 ratio betwen the four character classes, i.e. ( 432 + 23 + 14 + 459 ) ⁄ 4 = 928 ⁄ 4 = 232 in each category. We clearly do not have this, and have an excess of coloured smooth (C,Sh) and colourless wrinkled (c,sh) individuals. These must have the parental genotype, hence, the original cross was a coupling phase cross (C/C,Sh/Sh) X (c/c,sh/sh). The recombination frequency is ( 14 + 23 ) ⁄ 928 = 4%, indicating the genes are very closely linked.

    Kernel type

    Genotype

    Observed

    Expected

    χ2

    Coloured, smooth

    C/c, Sh/sh

    432

    232

    172

    Colourless, smooth

    c/c, Sh/sh

    23

    232

    188

    Coloured, wrinkled

    C/c, sh/sh

    14

    232

    204

    Colourless, wrinkled

    c/c, sh/sh

    459

    232

    222
    Σχ2 = 172 + 188 + 204 + 222 = 786. The critical value with 3 d.f. is 7.815, so our data are clearly nowhere even close to a 1:1:1:1 ratio.
  7. If intelligence were under highly-heritable polygenic control, two highly intelligent people's offspring would show regression to the population mean: although the children would stand a better chance of being brighter than the average person, they would be very unlikely to be brighter than their parents. Two averagely clever people would have about a 1 in 4 chance of producing children more intelligent than themselves, considerably more likely than the clever parents. Equally, there is a very good chance that the children of two very stupid people would be more intelligent than their parents.
  8. The number of legs a human being has a heritability essentially of zero. This may sound odd, but almost everyone with less than two legs has subnumery legs as the result of an accident or teratogen (thalidomide). Almost all variation in leg number is environmental, and not genetic. A 'gene' for something implies that there is a length of DNA that is associated with variation in a character, and which can be inherited by the next generation.
  9. The Y chromosome contains almost no genes, and all but one X chromosome in a cell is deactivated into a Barr body. Consequently XY, XX, XYY, XXY and X0 individuals all have essentially just a single complement of working X genes, plus-or-minus some Y genes. All other chromosomes are both active and full of genes. Consequently, almost all trisomies of autosomes are lethal. Unsurprisingly Down syndrome individuals have trisomy of the smallest autosome.
    • Observed carriers (Cf/cf) : 0.0296
    • Observed sufferers (cf/cf) : 0.0000093
    • Observed CF-negatives (Cf/Cf) : 1 − 0.0296 − 0.0000093 = 0.9703907
    • cf gene prevalence : 0.0000093 + ( 0.0296 ⁄ 2 ) = 0.0148093
    • Cf gene prevalance : 0.9703907 + ( 0.0296 ⁄ 2 ) = 0.9851907
    • Predicted carriers (Cf/cf) : 2 × 0.0148093 × 0.9851907 = 0.0292
    • Predicted sufferers (cf/cf) : 0.01480932 = 0.00022
    • Predicted CF-negatives (Cf/Cf) : 0.97039072 = 0.941658
    This population is not at HW equilibrium: there are about 20 times fewer suffers than we would expect, and correspondingly larger numbers of carriers and negatives. The most probable reason for this is selection: CF suffers have a life expectance of about 30, consequently, they make up a smaller proportion of the individuals. Interestingly, the actual proportion of live births for which the baby is a suffer is around 0.00026, very close to the predicted proportion, indicating that post-zygotic death rate is the overriding factor here (prezygotic factors, such as positive assortive mating between those negative for the cf gene, have little influence).
  11. Thymine is a legitimate base in DNA, so is much less detectable as a mutation than a deaminated cytosine, which forms uracil, a base that should not appear in DNA.