
These functions create and interact with
TRAMPknowns
objects (collections of known TRFLP
patterns). Knowns contrast with “samples” (see
TRAMPsamples
) in that knowns contain identified
profiles, while samples contain unidentified profiles. Knows must
have at most one peak per enzyme/primer combination (see Details).
TRAMPknowns(data, info, cluster.pars=list(), file.pat=NULL,
warn.factors=TRUE, ...)
# S3 method for TRAMPknowns
labels(object, ...)
# S3 method for TRAMPknowns
summary(object, include.info=FALSE, ...)
data.frame containing peak information.
data.frame, describing individual samples (see Details for definitions of both data.frames).
Parameters used when clustering the knowns database. See Details.
Optional partial filename in which to store knowns
database after modification. Files <file.pat>_info.csv
and <file.pat>_data.csv
will be created.
Logical: Should a warning be given if any columns
in info
or data
are converted into factors?
A TRAMPknowns
object.
Logical: Should the output be augmented with the
contents of the info
component of the TRAMPknowns
object?
TRAMPknowns
: Additional objects to incorportate
into a TRAMPknowns
object. Other methods: Further arguments
passed to or from other methods.
A new TRAMPknowns
object: a list with
components info
, data
(the provided data.frames, with
clustering information added to info
), cluster.pars
and file.pat
, plus any extra objects passed as ...
.
A sorted vector of the unique samples
present in x
(from info$knowns.pk
).
A data.frame, with the size of the peak (if
present) for each enzyme/primer combination, with each known
(indicated by knowns.pk
) as rows and each combination (in the
format <primer>_<enzyme>
) as columns.
The object has at least two components, which relate to each other (in
the sense of a relational database). info
holds information
about the individual samples, and data
holds information about
individual peaks (many of which may belong to a single sample).
Column definitions:
info
:
knowns.pk
:Unique positive integer, used to identify individual knowns (i.e. a “primary key”).
species
:Character, giving species name.
data
:
knowns.fk
:Positive integer, indicating which sample
the peak belongs to (by matching against info$knowns.pk
)
(i.e. a “foreign key”).
primer
:Character, giving the name of the primer used.
enzyme
:Character, giving the name of the restriction digest enzyme used.
size
:Numeric, giving size (in base pairs) of the peak.
In addition, TRAMPknowns
will create additional columns holding
clustering information (see group.knowns
). Additional
columns are allowed (and retained, but ignored) in both data.frames.
Additional objects are allowed as part of the TRAMPknowns
object, but these will not be written by
write.TRAMPknowns
; any extra objects passed (via
...
) will be included in the final TRAMPknowns
object.
The cluster.pars
argument controls how knowns will be clustered
(this will happen automatically as needed). Elements of the list
cluster.pars
may be any of the three arguments to
group.knowns
, and will be used as defaults in
subsequent calls to group.knowns
. If not provided, default
values are: dist.method="maximum"
,
hclust.method="complete"
, cut.height=2.5
(if only some
elements of cluster.pars
are provided, the remaining elements
default to the values above). To change values of clustering
parameters in an existing TRAMPknowns
object, use
group.knowns
.
A known contains at most one peak per enzyme/primer combination.
Where a species is known to have multiple TRFLP profiles, these should
be treated as separate knowns with different, unique, knowns.pk
values, but with identical species
values. A sample containing
either pattern will then be recorded as having that species present
(see group.knowns
).
TRAMPsamples
, which constructs an analagous object to
hold “samples” data.
plot.TRAMPknowns
, which creates a graphical
representation of the knowns data.
TRAMP
, for matching unknown TRFLP patterns to
TRAMPknowns
objects.
group.knowns
, which groups similar knowns (generally
called automatically).
add.known
and combine.TRAMPknowns
, which
provide tools for adding knowns from a sample data set and merging
knowns databases.
# NOT RUN {
## This example builds a TRAMPknowns object from completely artificial
## data:
## The info data.frame:
knowns.info <-
data.frame(knowns.pk=1:8,
species=rep(paste("Species", letters[1:5]), length=8))
knowns.info
## The data data.frame:
knowns.data <- expand.grid(knowns.fk=1:8,
primer=c("ITS1F", "ITS4"),
enzyme=c("BsuRI", "HpyCH4IV"))
knowns.data$size <- runif(nrow(knowns.data), min=40, max=800)
## Construct the TRAMPknowns object:
demo.knowns <- TRAMPknowns(knowns.data, knowns.info, warn.factors=FALSE)
## A plot of the pretend knowns:
plot(demo.knowns, cex=1, group.clusters=TRUE)
# }
Run the code above in your browser using DataLab