When comparing protein sequences between species, how does one determined what the degree of conservation is?
For example, is 87% identity considered conserved between mice and humans? How about yeast and humans? How about chimps and humans? I'm assuming this changes based on the evolutionary distance between species, but how does one calculate the threshold values (if there are any)?
Also, if I recall correctly there are two terms associated with conservation, identity and similarity (not sure about the latter). If someone can provide a technical definition of these terms that would be great.
If anyone knows of a review covering this topic, that would also be welcome.
Firstly, you can analyze conservation between nucleotide sequences or between protein sequences. Analyzing protein conservation you are mainly focused to functional constraints while nucleotide sequences maintain traceable some features related with their evolutionary origin. It’s also interesting to compare the level of nucleotide conservation with the aminoacid sequence conservation.
Conservation is a concept that usually is a relative concept and needs some reference conservation data to compare and to decide if your protein is more or less conserved that the set of proteins of your set (considered as a whole). If you are evaluating the level of conservation between pairs of species perhaps you could analyze the average conservation level between both species and then compare the conservation of each pair of ortholog proteins with the average conservation level. In this case it could be problematic evaluate the level of conservation for proteins without any ortholog in the other organism of the pair to be compared.
Another approach is the comparison of the conservation level between two ortholog proteins taking as reference the average level of conservation within the members of the protein family to which pertain the proteins to be compared. There are strong differences between the expected levels of conservation for different types of proteins. Thus, enzymes are very conserved even between taxonomically distant organism, proteins related with DNA physiology are also very conserved across distant genera,…
Depending on the goal of your study you can analyze conservation using different approaches.
I don't think that saying "Analyzing protein conservation you are mainly focused to functional constraints while nucleotide sequences maintain traceable some features related with their evolutionary origin" is true... Both protein and nucleotide sequences are used for studying evolutionary origins of proteins and organisms... Just one outstanding example: http://www.nature.com/nature/journal/v465/n7295/full/nature09014.html
There are a lot of important papers about evolutionary questions, including horizontal gene transfer, based on protein similarity because nucleotide sequence similarity implies protein similarity but the analysis of nucleotide sequences adds additional features (codon usage, k-tuple frequencies, GC percentage,..) that can delineate with more precision the origin of a sequence.
Yes yes, sorry, I just wanted to say that your statement could be wrongly interpreted as that proteins DO NOT "maintain traceable some features related with their evolutionary origin" contrary to the nucleotide sequences, which is of course not true ;-)
I don't think that saying "Analyzing protein conservation you are mainly focused to functional constraints while nucleotide sequences maintain traceable some features related with their evolutionary origin" is true... Both protein and nucleotide sequences are used for studying evolutionary origins of proteins and organisms... Just one outstanding example: http://www.nature.com/nature/journal/v465/n7295/full/nature09014.html
There are a lot of important papers about evolutionary questions, including horizontal gene transfer, based on protein similarity because nucleotide sequence similarity implies protein similarity but the analysis of nucleotide sequences adds additional features (codon usage, k-tuple frequencies, GC percentage,..) that can delineate with more precision the origin of a sequence.
Yes yes, sorry, I just wanted to say that your statement could be wrongly interpreted as that proteins DO NOT "maintain traceable some features related with their evolutionary origin" contrary to the nucleotide sequences, which is of course not true ;-)