Trends in Genetics
Volume 16, Issue 5, 1 May 2000, Pages 227-231
Journal home page for Trends in Genetics

Review
Homology: a personal view on some of the problems

https://doi.org/10.1016/S0168-9525(00)02005-9Get rights and content

Abstract

There are many problems relating to defining the terminology used to describe various biological relationships and getting agreement on which definitions are best. Here, I examine 15 terminological problems, all of which are current, and all of which relate to the usage of homology and its associated terms. I suggest a set of definitions that are intended to be totally consistent among themselves and also as consistent as possible with most current usage.

Section snippets

The other homologies problem

Organic chemists consider compounds such as methane, ethane and propane to be an homologous series because each differs from the next by a CH2 group. Thus, homoserine has one more CH2 group than serine. Mathematicians have special meanings for the term as well. There is no point in worrying about these differences, except to suggest that molecular biologists, mathematicians and bioinformaticists working in the field of biology learn and adhere to the biological definitions.

The redefinition problem

Homology was first defined in biology with something like its present meaning by Owen in 1843 who characterized homology as ‘the same organ under every variety of form and function’6. Common ancestry is not mentioned in that definition, which is unsurprising given that these were pre-Darwinian and pre-Mendelian times. Owen’s definition of homology emphasizes structure and location rather than ancestry. Some would have us return to Owen’s definition, perhaps out of a sense of precedence or some

The character/character-state problem

Many systematists, and nearly all molecular evolutionists, distinguish between a character, say amino acid, and its character states, say glycine and phenylalanine. This useful distinction is not universal. Many systematists will, if two character states are not the same, assert that the characters are non-homologous! This is confusing because it implies that the two characters do not have a common ancestor, which, if true, means they should not have been comparing the character states in the

The homology/homoplasy problem

Analogy describes characters whose similarity arises from convergent processes. Homology describes characters, irrespective of their character states, whose similarity arises after divergence from a common ancestral form. Homoplasy is the complement of analogy in that these two categories constitute all known non-random explanations of similarity. Some would make homoplasy, a term introduced by Lankester7, the complement of homology. But homoplasy is a relation of two character states in a

The recognition of homology problem

How does one know for sure that two sequences are homologous? One would always ‘know’ if we defined homology objectively as their having amino acids or nucleotides that are at least X% identical. But what is the appropriate value of X? One should not define homology objectively because: (a) it requires defining homology by an arbitrary amount of identity; (b) it excludes the possibility of analogy; and (c) this still does not solve the problem of our confidence that the characters asserted to

The homology subset problems

There are three disjoint subtypes of homology. Orthology is that relationship where sequence divergence follows speciation, that is, where the common ancestor of the two genes lies in the cenancestor of the taxa from which the two sequences were obtained9. This gives rise to a set of sequences whose true phylogeny is exactly the same as the true phylogeny of the organisms from which the sequences were obtained. Only orthologous sequences have this property.

Paralogy is defined as that condition

The gene loss problem

Imagine that a gene duplicated and then, following a subsequent speciation, one lineage lost one gene and the second lineage lost the other gene. What does one call the relationship of the remaining two genes? Paralogy, of course. The definition of the forms of homology does not change by virtue of the known, suspected, or unknown presence of a copy of a gene. This brings up the related problem, can a gap be a homolog? Yes. The alignment of molecules often needs indels (gaps required because

The structure/function problem

The definition of homology is about characters. Examples include genic (molecular), structural (morphological), functional (metabolic, regulatory, and behavioral) characters, and no doubt others. Should all characters be admitted? There are many examples in different organisms where the same structure has different functions or different structures have the same function. A marvelous example is the reptilian articular and quadrate bones of the mandible, which are orthologous to the mammalian

The bird/bat limbs problem

Are their forelimbs homologous or not? The forelimbs of the bat and the bird are adapted to flight, but the evolution to flight occurred independently in each lineage. Their cenancestral limb is the forelimb of a flightless reptile that is itself the reptilian cenancestor of the birds and mammals. Thus the limbs are (structurally) orthologous. On the other hand, the flight of birds and bats is (functionally) analogous.

The parallel/convergence problem

Five possible relationships for two changes in a character are shown in Fig. 2. They are given the names that generally conform to the English meanings of the words. However, many use the term convergence for changes that are parallel. Calling convergent that which does not converge can only be another source of confusion and should be resisted.

The homology/analogy problem

There is a tendency to assume that if two characters are significantly similar they must be homologous. This assumption has been proven to be untrue many times when the characters were morphological or behavioral. For nucleotide and amino acid sequences, the situation is different. Most of the time, the degree of similarity is so great that one (including me) will say that convergence could not have caused this much similarity17. It is commonly believed that there is no test that can

The gene conversion problem

Consider a gene duplication creating paralogs, followed some time later by a speciation event, and then by gene conversions in which copies of blocks of DNA from one gene simply replace the homologous residues in its paralog. This causes the paralogs to look more like each other. Indeed, they might look so much more alike than they otherwise should that, although they continue to look like paralogs within a species, they will appear to have duplicated recently and independently in each species,

The recombination problem

Two sequences or domains might have a common ancestor, in which case they are homologous, irrespective of the degree of similarity. A gene can be constructed from the domains of several other different genes. For example, enterokinase has at least five domains in addition to the protease domain24. One domain is related to a low-density lipoprotein receptor, another to a metalloprotease of the renal glomerulus, another to the Drosophila dorsal–ventral patterning gene and yet another to

The tandem repetitive characters problem

Some DNA is composed of chunks that are tandemly repeated, sometimes many times. The chunks are necessarily paralogous but they do represent a special case. Normally, the presence of gaps in an alignment is the result of a simple insertion or deletion (indel) in a gene. In the case of repeat arrays, the gaps can be the result of there being different numbers of the repeats. They should be treated differently from indels when building phylogenies but there is no nomenclatural problem; they are

The gene/allele problem

Are two alleles paralogs? Presumably not, given that the definition includes gene duplication. But consider the following. The earliest vertebrates had only one hemoglobin gene and so there was no cooperativity that allowed the more efficient transport of oxygen. Then a mutation arose that permitted some slight but beneficial degree of heterozygous cooperativity. It would be selected for until perhaps the population had about equal frequencies of the two alleles. At this point, only half the

Conclusion

I recognize, and even accept, that homology has been used by various people with different meanings, even though similarity was a common denominator among these meanings. The two most important of these meanings related homology to similar structures and/or to similar functions. (By structures I mean both molecular sequences and morphology.) Life would have been simple had phylogenetic homology necessarily implied structural homology or either of them necessarily implied functional homology.

References (25)

  • R. Owen (1843) Lectures on the comparative anatomy and physiology of the invertebrate animals. Longman, Brown, Green &...
  • E.R. Lankester

    On the use of the term homology in modern zoology

    Ann. Mag. Nat. Hist. Ser.

    (1870)
  • Cited by (455)

    • Quasi-best match graphs

      2023, Discrete Applied Mathematics
    View all citing articles on Scopus
    View full text