Indeed. The Taylor Venn Diagram is not based on emprical scales of "hydrophobicity" etc, but on observed substitution data. This captures a lot more detail and subtlety from evolutionary constraints on protein structure than biophysical measurements that only consider one property.
Geoff. On 02/11/2016 10:10, James Procter wrote: > Hi. > > Interesting question Mungo - and perhaps you may not have realised that > it doesn't have a clear cut answer. > > On 02/11/2016 09:11, Mungo Carstairs (Staff) wrote: >> The question was whether Threonine should be classed as hydrophobic (in >> which case the diagram needs amended) or not (in which case Jalview code >> needs amended). > Threonine does have some hydrophobic characteristics - but these are > context dependent. The sidechain is relatively small, and dominated by > polar groups, so whilst it doesn't hydrogen bond terribly well with > water, its polarizable nature allows it to mix, so the entropic > solvation cost is not huge. > > The original reference for this diagram is Taylor 1986 'The > classification of amino acid conservation', and many people have redrawn > it in subsequent publications. Most interpretations of physicochemical > properties at that time were based on empirical scales derived from > solubility measurements of the monomer, rather than any protein > environment based rational (although there were sufficient structures > available to support the hydrophobic core theory). As we get more > structures, we also get a better idea of amino acid spatial preferences. > >> Question remains, which is it (for Jalview conservation purposes at least)? > What is important for Jalview right now is that the method and reference > data used are those that is widely accepted (warts and all). > > I think it would be posible to offer configuration to support different > forms of the matrix in a future release, but only experts will really > use that feature - and then, they may prefer to use entirely different > forms of physichemical conservation measurement. I've mentioned one key > resource for this before: http://www.genome.jp/aaindex/ > > For further reading I suggest you first take a look at Valdar 2002 : > valdarlab.unc.edu/papers/proteins2002.pdf > > > Jim. > -- Geoff Barton | Professor of Bioinformatics | Head of Division of Computational Biology School of Life Sciences | University of Dundee, Scotland, UK | [email protected] Tel: +44 1382 385860 | www.compbio.dundee.ac.uk | twitter: @gjbarton The University of Dundee is registered Scottish charity: No.SC015096 The University of Dundee is a registered Scottish Charity, No: SC015096 _______________________________________________ Jalview-dev mailing list [email protected] http://www.compbio.dundee.ac.uk/mailman/listinfo/jalview-dev
