If DNA samples match exactly, they identify an individual. What is the percentage for other relationships?

For each of the biological classifications, what is the range of DNA matches: parent, sibling, other family, species, genus, family, order, class, phylum?If DNA samples match exactly, they identify an individual. What is the percentage for other relationships?
The only one of these for which an exact number can be determined is a parental relationship. Due to the way sex cells form and then unite during fertilisation, a child receives one set of chromosomes from each parent, meaning an 50% match to each. Which is actually 50% of the DNA that varies across the population at all, since in reality all human genomes are more than 99% identical to each other! Different alleles of a gene may differ only in one single base pair - see sickle cell anaemia for an example -, so even if you've got two different things from each parent, these ';different'; things might well be 99.9% identical!





Siblings ON AVERAGE are 50% similar (in terms of that small section of DNA that carries variation, see above), but there is a (very small) chance that they get just the other half of each parent's genome, and they won't be more related genetically than any two members of the population.





For species? Well, this varies greatly between species, and depends on population size in the past and present, number of more or less isolated populations, time the species has been around, etc etc. Classification into species is not based on genetic distance but other, equally controversial attributes, like the ability to produce fertile offspring.





As for even larger categories of classification, even distant relatives can be surprisingly similar. However, different organisms don't have the same amount of DNA, so in distantly related creatures it's not very meaningful to check the whole genome, I guess. There are some genes and proteins found in just about every organism - like the protein cytochrome c - which can be used to determine relationships. The surprising thing is that cytochrome c of a mammal - like a horse, or a human - is about 60% identical to that of a plant, and still 40% identical to that of a bacterium... The degree of similarity depends on how mutations affect the functioning of a protein - if they don't make the protein useless, then the genes of a mutant protein are more likely to be passed on, so more mutations are accumulated through generations.