What Classical Genetics Says About
When diploid organisms acquire a new gene by transformation, often a single copy is picked up, but sometimes two or more. It is now known that the transforming genes are integrated randomly in the recipient chromosome, but it is not inconceivable (although extraordinarily unlikely, given the size of eukaryotic genomes) that a [correcting] gene, such as claimed in Ledoux's results, would simply replace the mutant locus or be very closely linked to it. Classical genetics can help unravel some different possibilities.
Following are four simplified examples that are applicable to Arabidopsis transformation claims, which we now know were artifactual. In these examples, I will use hypothetical chromosomes named X, Y, and Z (without reference to sex chromosomes) and I will call a a recessive lethal mutation and A* a dominant allele contributed by foreign DNA taken up by the organism. The recipient lethal mutants are homozygous recessive for the a locus on chromosome Z and are thus ZaZa. Chromosomes X and Y do not carry the a locus and could be any other chromosome, except Z.
and Integrated Randomly in Chromosome X
In this case, the organism has the genotype XXA*ZaZa, where X denotes the regular chromosome X, and XA* represents the other X, having acquired a copy of A* by transformation. The gametes produced by this individual are