Alignment and phylogenetic
analysis
The two commonly used methods for phylogenetic analysis are distance
matrix and maximum parsimony. Distance matrix methods compare
all of the possible pairs of sequences in an alignment and generate
scores based on their similarity in order to construct a matrix
which is then used to progressively align the sequences, starting
with the most similar. Maximum parsimony methods assume that the
evolutionary path from a common ancestor to the current diversity
is the simplest possible one and construct a tree which requires
the minimum number of mutations. In this study, because of the
large size of the data set, a distance matrix analysis was performed
with the ClustalW package [1]. Maximum parsimony
analysis was found to yield similar results to distance matrix
analysis in earlier studies on a smaller data set [2].
Maximum parsimony calculations, however, are particularly intensive
and can give artefactual groupings of sequences especially at
the end of long branches [3].
A preliminary (pairwise) alignment of all of the full length sequences was made with a modified version of the method of Feng and Doolittle in ClustalW [4]. The sequences were then trimmed to produce a core motor domain. This core motor domain corresponds to residues 88 to 780 of chicken skeletal muscle myosin II from which the first crystal structure of a myosin head was solved [5]. A second guide alignment with these core motor domain sequences was produced, followed by a final multiple alignment with correction for multiple substitutions [6]. Optimally, when producing a phylogenetic tree from such an alignment, any positions in the alignment with gaps are ignored. Such a strategy would have excluded half of the data because of the large number of sequences, thus positions with gaps were included. An unrooted phylogenetic tree was constructed and tested for branching order confidence by bootstrapping [7]. Bootstrapping redraws 1000 trees with random samples from the multiple alignment. If the same branches were found at a certain node in >900 out of those 1000 random trees, then that node was accepted as being significant and therefore the branching order at that point in the tree was well supported. (Bootstrapping data)
Further manipulation of the
alignment and preparation for publication.
The alignment itself was exported from ClustalW in .pir format
and read into SeqVu (Shareware alignment programme, ©Garvan
Institute 1992-1995. Author: James
Gardner). The alignment was examined by eye and optimised
for publication, with residues conserved in >90% of the 82
myosins boxed and all residues coloured according to the following
scheme: non-polar (A,I,L,V,P,F,W,M) - green; uncharged polar (S,T,Y,C,N,Q)
and glycine - yellow; Acidic (D,E) - red; Basic (H,K,R) - blue.
While a useful alignment refinement programme, SeqVu is limited
in output options and it was decided that the best way of getting
the pages from it into a useful graphics programme was by using
a Mac extension called PictPocket This allows windows to be grabbed
into the clipboard as pict items, with each graphical constituent
as an independently selectable item. This was pasted into ClarisDraw
on a PowerMac 9500 (the presence of so many pict items required
a high-speed mac) Additional graphical features (secondary structure,
etc) were added in ClarisDraw. Precise secondary structure assignments
were decided using a Silicon Graphics Indigo II workstation, running.
Quanta software (MSI Inc.)
References
[1] Rayment, I., Rypniewski, W.R., Schmidt-BÑse, K., Smith, R., Tomchick, D.R., Benning, M.M., Winkelmann, D.A., Wesenberg, G. and Holden, H.M. (1993) Three-dimensional structure of myosin subfragment-1: A molecular motor. Science, 261:50-58.
[2] Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) Clustal-W - improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research, 22(22):4673-4680.
[3] Goodson, H.V. (1994) Molecular evolution of the myosin superfamily: application of phylogenetic techniques to cell biological questions, in Molecular Evolution of Physiological Processes. The Rockefeller University Press: New York. pp. 141-157.
[4] Felsenstein, J. (1988) Annual Review of Genetics, 22:521-565.
[5] Feng, D.F. and Doolittle, R.F. (1987) Progressive sequence alignment as a prerequisite to correct phylogenetic trees. Journal Of Molecular Evolution, 25(4):351-360.
[6] Fisher, A.J., Smith, C.A., Thoden, J.B., Smith, R., Sutoh, K., Holden, H.M. and Rayment, I. (1995) X-Ray structures of the myosin motor domain of Dictyostelium discoideum complexed with MgADP-BeFx and MgADP- AlF4-. Biochemistry, 34(28):8960-8972.
[7] Kimura, M., The Neutral Theory of Molecular Evolution. 1983, Cambridge: Cambridge University Press.
[8] Felsenstein, J. (1985) Confidence limits on phylogenies - an approach using the bootstrap. Evolution, 39(4):783-791.
The Myosin Site and Alignment Page initiated and produced by Jamie Cope and subsequently updated by Tony Hodge to whom comments should be addressed.
Medical Research Council, Laboratory of
Molecular Biology,
Hills Road,
Cambridge
UK
CB2 2QH
If you use diagrams, trees, or sequence
alignments from the Myosin site, we ask that you cite either the
Home Page and authors, or the appropriate source publication in
your work.
Copyright 2000 All rights reserved
Last updated Tuesday January 25, 2000