Color residues by a numeric value computed in R, such as a
conservation score, a mutation effect or a p-value. The values are mapped
onto a color gradient, and the result can be used as color wherever
a representation takes one: addRepresentation,
addSelection, updateColor and
updateRepresentation.
colorByValue(
values,
palette = c("blue", "white", "red"),
domain = NULL,
na_color = "grey"
)Either a data frame with columns resno and
value, and optionally chain, inscode (insertion code)
and structure (the structure number, counting from 1); or a named
numeric vector with names "chain:resno", such as "A:12",
"A:12B" (with insertion code B) or "12" (any chain).
Without chain a row applies to that residue number in every chain,
and without structure to every structure. The data frame of the
Shiny input input$<outputId>_residues with an added value
column can be used as is.
Two or more colors (names or HEX codes) for the gradient, from the lowest to the highest value.
The values at the two ends of the palette, c(min, max).
Defaults to the range of values. Values outside it get the color of
the nearest end.
Color for residues with a missing value (NA) and for
residues not listed in values.
A NGLVieweR_colorByValue object to use as color.
The colors are computed in R, so the viewer shows exactly the
colors that grDevices::colorRamp(palette) gives. A row for a
specific chain or structure takes precedence over a row for any chain or
structure. NGL.js cannot show a color legend; the returned object
contains palette and domain (as HEX colors and numbers) so a
legend can be drawn next to the viewer, e.g. with a plot.
scores <- data.frame(chain = "A", resno = 1:46, value = sin(1:46 / 5))
NGLVieweR("1CRN") %>%
addRepresentation("cartoon",
param = list(color = colorByValue(scores,
palette = c("blue", "white", "red"),
domain = c(-1, 1))))
# A named vector, for a few residues; all others are grey
NGLVieweR("1CRN") %>%
addRepresentation("cartoon",
param = list(color = colorByValue(c("A:10" = 1, "A:20" = 2, "A:30" = 3))))
if (interactive()) {
library(shiny)
ui <- fluidPage(
sliderInput("k", "Score = sin(resno / k)", 1, 20, 5),
NGLVieweROutput("structure")
)
server <- function(input, output) {
output$structure <- renderNGLVieweR({
NGLVieweR("1CRN") %>%
addRepresentation("cartoon", param = list(name = "cartoon"))
})
observeEvent(input$k, {
residues <- input$structure_residues
req(residues)
residues$value <- sin(residues$resno / input$k)
NGLVieweR_proxy("structure") %>%
updateColor("cartoon", colorByValue(residues, domain = c(-1, 1)))
})
}
shinyApp(ui, server)
}