This page gives access to the supplementary material belonging to the paper:
Creation of a multiple alignment
A multiple alignment containing 43 ATP- and PPi-dependent PFK
sequences was created using the program ClustalW. The alignment
was adjusted manually by using information on the secondary structure
of the E. coli ATP-PFK. From this alignment a sub-collection
of relevant sequences was made as shown in Fig.
3 in the paper.
Phylogenetic analyses.
From the combined alignment of all
available ATP-dependent and PPi-dependent PFKs, those regions that
contained major insertions or deletions, or those where the alignment
was ambiguous, were removed). The following columns in the alignment
were retained:
271,296:310,351:375,388:392,446:450,479:498,505:509,522:602,642. The
resulting condensed alignment file in
Phylip format containing the remaining 228 sites was used in
phylogenetic analyses with either the neighbor-joining, maximum
parsimony, or the maximum likelihood method. All three methods gave
highly similar results and the neighbor-joining
and maximum likelihood trees are
available. The clustering into separate clades of the eukaryotic
ATP-PFKs (N- and C-terminal domains), the bacterial ATP-dependent
PFKs, as well as a all PPi-related PFKs, was supported by
bootstrapping analysis for both methods applied. Slight differences
between the methods were observed. Maximum likelihoodwas not able to
resolve the relative branching order within some of the clades such
as the bacterial ATP-PFK clade. Neither did it resolve the exact
order of the separation between bacterial and eukaryotic ATP-PFKs and
the gene duplication in the latter. In the neighbor-joining method
Propionobacterium. freudenreichii branched off before the
separation of N. fowleri, E. histolitica and T. brucei,
whereas in the ML method it grouped together with the three protists
sequences. Under maximum parsimony the three protist PFKs formed one
clade (not shown). Also in the ML two the helminths, S.
mansoni and H. contortus, were monophyletic, while with
neighbor joining they behaved a paraphyletic taxa. A distance matrix
was prepared after conversion of the alignment file in genbbank
format to a pir/nbrf format followed by importation into ClustalW.
The Uncorrected distance matrix was
prepared with the settings "correction for multiple mutations" = off
and "exclusion of regions with gaps" = off.
Access the directory with all avaialable files