After running the registration function register()
as
shown in the Registering
data article, users can summarise and visualise the results as
illustrated in the figure below.
The total number of registered and non-registered genes can be
obtained by running the function summarise_registration()
with registration_results
object as an input.
The function summarise_registration()
returns a list
containing three different objects:
summary
is a data frame containing the summary of the
registration results,registered_genes
is a vector of gene accessions or IDs
which were successfully registered, andnon_registered_genes
is a vector of non-registered gene
accessions or IDs.# Get registration summary
reg_summary <- summarise_registration(registration_results)
reg_summary$summary |>
knitr::kable()
Result | Value |
---|---|
Total genes | 10 |
Registered genes | 9 |
Non-registered genes | 1 |
Stretch | [2.25, 4] |
Shift | [-6.24, 4.4] |
The list of gene accessions which were registered or not registered can be viewed by calling:
The function plot_registration_results()
allows users to
plot the registration results of the genes of interest.
Notice that the plot includes a label indicating if the particular genes are registered or not registered, as well as the registration parameters in case the registration was successful.
For more details on the other function paramaters, go to
plot_registration_results()
.
After registering the data, users can compare the overall similarity
between datasets before and after registering using the function
calculate_distance()
.
The function calculate_distance()
returns a list of two
data frames:
registered
distance between scaled reference and query
expressions using registered time points.original
distance between scaled reference and query
expressions using original time points.Each of these data frames above can be visualised using the
plot_heatmap()
function, by selecting either
type = "registered"
or type = "original"
.
# Plot heatmap of mean expression profiles distance before registration process
plot_heatmap(
sample_distance,
type = "original"
)
# Plot heatmap of mean expression profiles distance after registration process
plot_heatmap(
sample_distance,
type = "registered",
match_timepoints = TRUE
)
Notice that we use match_timepoints = TRUE
to match the
registered query time points to the reference time points.