|
|
Basic concepts and vocabulary
Mesquite was designed to explore the evolution of organsims and
their characteristics. Here is some basic vocabulary that relates
the biology to the implementation in Mesquite.
Taxa
- Taxon - One of the basic entities being studied, usually
a species or a gene sequence. In Mesquite (as in other programs),
the word "Taxon" has shifted from its traditional
meaning to become more or less equivalent to "terminal
taxon" (the smallest unit of analysis of relationships).
- Block of taxa - A collection of taxa grouped together
for analysis. In most programs dealing with phylogeny, there
is only a single block of taxa allowed in a file (e.g., "Primates",
consisting of 20 species of primates). In Mesquite, multiple
blocks are allowed (perhaps to study associations among them,
in the style of Rod Page's programs), and so it is useful to
refer to different blocks of taxa. A block of taxa is equivalent
to a TAXA block in a NEXUS file.
- Group Membership (Taxa partition) - A subdivision of
a taxa block into disjoint subsets. Thus, a taxa block for Mammals
may have its taxa partitioned into one of three groups, Carnivore,
Ominvore and Herbivore. A taxa partition is equivalent to a
TAXAPARTITION in a NEXUS file.
Characters
- Character - A variable or characteristic measured or
observed for a block of taxa. It could be from molecular data
("Nucleotide at position 367 of the 18S gene"), or
it could be phenotypic ("length of wing"). Characters
in Mesquite typically exist only within character matrices.
- Character model - a model of the evolution of a character.
This could be a model for use in parsimony calculations (e.g., "unordered",
a cost matrix) or in likelihood calculations or stochastic
simulations (e.g., HKY85, GTR).
- Character matrix - A matrix representing the character
states of various characters for a block of taxa. Each character
matrix in Mesquite must be homogeneous with respect to type
of character (e.g. nucleotide sequences can't be intermingled
with morphometrics data).
- Character partition - A subdivision of a characters
into disjoint subsets. Thus, a character matrix for insects
may have its characters partitioned into one of three groups,
larval, pupal, and adult. A character partition is equivalent
to a CHARPARTITION in a NEXUS file.
- Character inclusion set - A specification of what characters
are to be included and what excluded from analyses. This is
equivalent (in inverse form) to an EXSET in a NEXUS file.
- Character weight set - A specification of what weights
are to be applied to characters in summed calculations (especially
for parsimony calculations). This is equivalent to a WTSET in
a NEXUS file.
- Parsimony model set - A specification of what parsimony
model of character evolution (i.e.,
what
"transformation type") is to be applied to each character
of a matrix. This is equivalent to a TYPESET in a NEXUS file.
- Probability model set - A specification of what probalistic
model of character evolution is
to be
applied
to each character of a matrix.
Trees
- Tree - A tree representing the relationships of the
taxa. If the taxa are species, the tree represents phylogeny;
if the taxa are gene sequences, the tree represents a gene genealogy.
- Block of trees - A collection of trees. This might
be a set of trees saved by the user, or output by a program
like PAUP*.
|
|
