smear:
Plots log-ratio transformed abundances pairwise.
Index-aware, meaning that it only plots pairs indexed
in @pairs, unless no pairs are indexed.
dendrogram:
Plots a clustering of proportionality matrix.
Index-aware, meaning that it only plots pairs indexed
in @pairs, unless no pairs are indexed.
Heatmap intensity is not scaled.
slate:
Builds a table of VLR, VLS, and proportionality
for each feature pair in a propr object. If the
argument k is provided, the table will also
include co-cluster membership.
prism:
Plots the variance of the ratio of the log-ratio transformed
feature pair (VLR) versus the sum of the individual variances
of each log-ratio transformed feature (VLS). The ratio of
the VLR to the VLS equals 1 - rho. As such, we use
here seven rainbow colored lines to indicate where rho
equals [.01, .05, .50, 0, 1.50, 1.95, 1.99], going
from red to violet.
bokeh:
Plots the feature variances for each log-ratio transformed
feature pair in the propr object. Highly proportional
pairs will aggregate near the y = x diagonal.
Clusters that appear toward the top-right of the
figure contain features with highly variable abundance across
all samples. Clusters that appear toward the
bottom-left of the figure contain features with fixed
abundance across all samples. Uses a log scale.
bucket:
Plots an estimation of the degree to which a feature pair
differentiates the experimental groups versus the
measure of the proportionality between that feature pair.
The discrimination score is defined as the negative
log of the p-values for each feature in the pair,
computed independently using kruskal.test.
"It's pronounced, 'bouquet'." - Hyacinth Bucket
pca:
Plots the first two principal components as calculated
using the log-ratio transformed feature vectors. This
provides a statistically valid alternative to
conventional principal components analysis (PCA).
For more information, see <DOI:10.1139/cjm-2015-0821>.
snapshot:
Plots the log-ratio transformed feature abundance as
a heatmap, along with the respective dendrograms.
Heatmap intensity is not scaled.
cytescape:
Builds a table of indexed pairs and their proportionality.
In doing so, this function displays a preview of the
interaction network, built using igraph.
We recommend using the result as input to a
network visualization tool like Cytoscape.
smear(rho, prompt = TRUE, plotly = FALSE)dendrogram(rho, prompt = TRUE, plotly = FALSE)
slate(rho, k, prompt = TRUE, plotly = FALSE)
bucket(rho, group, k, prompt = TRUE, plotly = FALSE)
prism(rho, k, prompt = TRUE, plotly = FALSE)
bokeh(rho, k, prompt = TRUE, plotly = FALSE)
pca(rho, group, prompt = TRUE, plotly = FALSE)
snapshot(rho, prompt = TRUE, plotly = FALSE)
cytescape(object, col1, col2, prompt = TRUE, d3 = FALSE)
A logical scalar. Set to FALSE to disable
the courtesy prompt when working with big data.
A logical scalar. Set to TRUE to produce
a dynamic plot using the plotly package.
A numeric scalar. The number of clusters. Optional
parameter for bucket, prism, and bokeh.
Providing the argument k will color feature pairs
by co-cluster membership. In other words, a feature pair
will receive a color if and only if both features belong
to same the cluster (calculated using hclust).
A character vector. Group or sub-group memberships,
ordered according to the row names in @counts and
@logratio. Required parameter for bucket
and optional parameter for pca.
A propr object created from perb.
However, smear, dendrogram, and cytescape will
also accommodate results from phit and phis.
A character vector. Specifies which nodes
to color red or blue, respectively.
A boolean. Use rgl to plot 3D network.
smear, pca:
Returns a ggplot object.
dendrogram, snapshot:
Returns a dendrogram object.
slate:
Returns a data.frame of all pairwise relationships.
If the argument k is provided, returns a list of
the data.frame of pairwise relationships and the
cluster membership.
prism, bokeh, bucket:
Returns cluster membership if k is provided.
Otherwise, returns a ggplot object.
cytescape:
Returns a data.frame of indexed pairs.