Peter,
This is one of the most sensible questions to ask before begining a proje=
ct
and unfortunately, not enough people bother to ask it. Rather, there is =
a
herd-instinct to follow whatever has gone before.
First of all, in order to compare your data with previously-published dat=
a, it
might be desirable to sequence one of the 'popular' genes. These are the=
ribosomal RNA genes small and large subunit (and also the mitochondrials)=
,
elongation factors, rubisco, ATPases, GroESL genes, RecA, Cytochrome B, a=
nd more.
A huge number of genes are siutable for phylogenetic studies, but perhaps=
it
is sensible to sequence a gene for which there is a large database.
I would then suggest that you perform a preliminary study on a small
(judiciously chosen so that they are representative) number of taxa. Ali=
gn
the sequences and see if:
They are easily and unambiguously alignable.
There is sufficient variation among the sequences.
The resulting trees make sense (duplications and lateral transfers can
confound =
phylogeny reconstruction).
It would also be nice if all the variation is not concentrated on a few
residues. This could lead to the situation where the variable sites are
saturated with substitutions.
It will be interesting to hear other peoples thoughts on this. I don't t=
hink
that enough people spend enough time thinking about this issue.
My =A30.02 - worth.
James
Peter Schuchert wrote:
> =
> How would you select the most suitable gene for investigating
> the molecular phylogeny of a particular clade?
> =
> Peter
-- =
=3D=3D=3D=3D=3D=3D=3D=3D=3D
James O. McInerney email: J.mcinerney at nhm.ac.uk
Department of Zoology, phone: +44 171 938 9247
The Natural History Museum, Fax: +44 171 938 9158
Cromwell Road, =
London SW7 5BD. =
=3D=3D=3D=3D=3D=3D=3D=3D=3D
