Hi, @BradyAJohnston @swsoyee

About the m_add_res_labels()
Can you add some function to make the label scatter and with less overlap with the plot?
like this figure (probably made by AI or something like this)

Yang

• addArrow(spec) → {$3Dmol.GLShape}
• addAsOneMolecule(data, format) → {$3Dmol.GLModel}
• addBox(spec) → {$3Dmol.GLShape}
• addCurve(spec) → {$3Dmol.GLShape}
• addCustom(spec) → {$3Dmol.GLShape}
• addCylinder(spec) → {$3Dmol.GLShape}
• addIsosurface(data, spec) → {$3Dmol.GLShape}
• addLabel(text, options, sel, noshow) → {$3Dmol.Label}
• addLine(spec) → {$3Dmol.GLShape}
• addMesh(mesh, style) → {Number}
• addModel(data, format, options) → {$3Dmol.GLModel}
• addModels(data, format) → {Array.<$3Dmol.GLModel>}
• addModelsAsFrames(data, format) → {$3Dmol.GLModel}
• addPropertyLabels(prop, sel, style)
• addResLabels(sel, style, byframe)
• addShape(shapeSpec) → {$3Dmol.GLShape}
• addSphere(spec) → {$3Dmol.GLShape}
• addStyle(sel, style)
• addSurface(type, style, atomsel, allsel, focus, surfacecallback) → {Promise}
• addUnitCell(model, spec)
• addVolumetricData(data, format, or) → {$3Dmol.GLShape} deprecated Use addIsosurface instead
• addVolumetricRender(data, spec) → {$3Dmol.GLShape}
• animate(options)
• center(sel, animationDuration, fixedPath)
• clear()
• createModelFrom(sel, extract) → {$3Dmol.GLModel}
• enableFog(fog)
• exportJSON(includeStyles, modelID) → {string}
• exportVRML()
• fitSlab(sel)
• getFrame()
• getInternalState()
• getModel(id) → {GLModel}
• getNumFrames() → {number}
• getPerceivedDistance() → {number}
• getView() → {Array.}
• isAnimated() → {boolean}
• linkViewer(otherview)
• mapAtomProperties(props,, sel)
• pdbData(sel) → {string} (Same feature done by using {bio3d} #6 )
• pngURI() (#27)
• removeAllLabels()
• removeAllModels()
• removeAllShapes()
• removeAllSurfaces()
• removeLabel(label)
• removeModel(model)
• removeShape(shape)
• removeSurface(surf)
• removeUnitCell(model)
• render()
• replicateUnitCell(A, B, C, model)
• resize()
• rotate(angle, axis, animationDuration, fixedPath)
• selectedAtoms(sel) → {Array.}
• setAutoEyeSeparation() → {number}
• setBackgroundColor(hex, a)
• setClickable(sel, clickable, callback)
• setColorByElement(sel, colors)
• setColorByProperty(sel, prop, scheme)
• setContainer(element)
• setDefaultCartoonQuality()
• setFrame(framenum) → {Promise}
• setHeight(h)
• setHoverable(sel, hoverable, hover_callback, unhover_callback)
• setHoverDuration(hoverDuration)
• setInternalState()
• setLabelStyle(label, stylespec) → {$3Dmol.Label}
• setLabelText(label, text) → {$3Dmol.Label}
• setPerceivedDistance()
• setProjection()
• setSlab()
• setStateChangeCallback()
• setStyle(sel, style)
• setSurfaceMaterialStyle(surf, style)
• setView(arg)
• setViewChangeCallback()
• setViewStyle()
• setWidth(w)
• setZoomLimits()
• spin(axis)
• stopAnimate()
• translate(x, y, animationDuration, fixedPath)
• translateScene(x, y, animationDuration, fixedPath)
• vibrate(numFrames, amplitude, bothWays, arrowSpec)
• zoom(factor, animationDuration, fixedPath)
• zoomTo(sel, animationDuration, fixedPath)

A test for list converting is not difficult so adding a test would be prefer.
https://testthat.r-lib.org/index.html

Originally posted by @swsoyee in #6 (comment)

Hi, Su

How to set certain amino acids with a special color?
See this demo trimer protein with 6 chains, including HA1: A, C, and E; HA2: B, D, and F.
My purpose is set the 206:220 on HA1 with red color (#e41a1c).
So, I try to use m_add_style and m_add_surface. But they look like do not as work as my desire.
Is there something wrong with my code?

rm(list=ls()); gc()
library(bio3d)
library(r3dmol)

p <- r3dmol() %>%
  m_add_model(data = m_fetch_pdb("6D7U", save.pdb = TRUE)) %>%
  m_add_model(data = m_bio3d(pdb)) %>%
#   m_set_style(style = m_style_cartoon(color = 'spectrum', ribbon = TRUE)) %>%
  m_set_style(style = m_style_cartoon()) %>% 
  m_add_outline() %>% 
#   m_set_style(style = c(m_style_cartoon(color = "spectrum", ribbon = TRUE), m_style_line())) %>% 
  m_add_surface(atomsel = m_sel(chain =c('A', 'B')), style = m_style_surface(opacity = 0.9, colorScheme="chain")) %>% 
  m_add_style(sel = m_sel(resi = c(206:220), chain="E"), style = m_style_sphere(scale = 0.5, color = "#e41a1c")) %>%
#   m_set_style(sel = m_sel(resi = c(206:220), chain = "A", byres = TRUE), style = m_style_cartoon(color = "red")) %>%
  m_add_surface(atomsel = m_sel(resi = c(206:220), chain = "C"), style = m_style_surface(opacity = 0.9, color = "#e41a1c")) %>%
#   m_add_line(start = m_sel(resi = 1:10, chain = "A"), end = m_sel(resi = 1:10, chain = "B"), spec = list(dashed = TRUE)) %>%
  m_add_label(text = "Site A", sel = m_sel( resi = 206:220, chain = "C"), style = m_style_label(borderColor = "green", borderThickness = 1, inFront = FALSE)) %>%
#   m_zoom_to(sel = m_sel(resi = c(206:220), chain = "A"))
#   m_set_zoom_limits(upper = 200, lower = 30)
    m_zoom_to()
p

Thanks,
Yang

https://github.com/3dmol/3Dmol.js/releases/tag/1.7.0
If anyone can help it would be greatly appreciated.

在 Rstudio 里可以输出图片，但是在 jupyter Lab 里无法输出。

不知道是不是哪里不对？

I have noticed that if I don't use m_zoom_to() then parts of the structure disappear as I rotate. I would like to be able to turn labels on and off in a Shiny app so I don't want it to reset the zoom level each time I change something.

model=r3dmol(                         # Set up the initial viewer
      viewer_spec = m_viewer_spec(
        cartoonQuality = 10,
        lowerZoomLimit = 50,
        upperZoomLimit = 1000
      )
    ) %>%
    m_add_model(                 # Add model to scene
        data = m_bio3d(bio3d::read.pdb("https://alphafold.ebi.ac.uk/files/AF-Q9Y6K1-F1-model_v4.pdb")),
        format = "pdb"
    )

I predicted the 3D structure of my protein, I am wondering whether I can input the generated pdb data?

Hi @swsoyee ,
I have been trying to export the image generated through m_png() through a download button in Shiny, but I am not able to get that work. I have tried png-dev.off(), png::writePNG and also htmlWidgets-webshot but none of them worked for me. They work normally in an R environment but not in a shiny app download button. Here is the corresponding code from the server section of my app.R:

expression <- {
  r3dmol( viewer_spec = m_viewer_spec(cartoonQuality = 10, lowerZoomLimit = 50, upperZoomLimit = 350) ) %>%
    # Add model to scene
    m_add_model(data = pdb_file, format = "pdb") %>%
    # Zoom to encompass the whole scene
    m_zoom_to() %>%
    # Set style of structures
    m_set_style(style = m_style_cartoon(color = "#00cc96")) %>%
    # Set style of specific selection (selecting by secondary)
    m_set_style( sel = m_sel(ss = "s"), style = m_style_cartoon(color = "#636efa", arrows = TRUE) ) %>%
    # Style the alpha helix
    m_set_style( sel = m_sel(ss = "h"), style = m_style_cartoon(color = "#ff7f0e") ) 
}

output$pdb <- renderR3dmol(expression)

data2 <- reactiveValues()

data2$imageToSave <- renderImage({
  expression %>% m_png()
  })

output$downloadImage <- downloadHandler(
  filename = "AuxPhosImage.png",
  content = function(file){ 
    png::writePNG(data2$imageToSave)
  }
  )

I am using 'pdb' and 'downloadImage' in the 'ui' part as fluidRow(column(6, align="center", downloadButton('downloadImage','Download Image'), r3dmolOutput("pdb")).

I tried the same procedure for to save a ggplot object using ggsave which worked perfectly. But I am not able to get this working. One difference that I found is that this png from r3dmol is rendered into the 'viewer' but not the 'plot' section, for which the reason the png saving options are not working I guess, I might be wrong.

Any help is highly appreciated.

Thanks, Sumanth.

Support for creating multiple lines at once.

Currently, the m_add_line() will only create a single line from the two given points in space. Ideally I would like to be able to feed a vector / list of points into the start and end arguments, and it will create multiple lines.

I've tried playing around with looping over the m_add_line() multiple times, but it doesn't work properly. I don't know enough about the javascript side of things as to why it's not working, but would appreciate ideas as to how to get it to work.

Potential also for applying to m_add_cylinder(), m_add_sphere() etc

Thank you for such a useful package -- I've found it much easier to use than some of the alternatives.

m_add_label prints a label, but if given m_vector3 coordinates, the label prints at the wrong position.

The following reprex shows this, using the example code from the help file. Here, I print 2 labels, at different positions (I just changed the signs of the coordinates and label colors) but the labels end up in the same position.

library(r3dmol)

r3dmol() %>%
  m_add_model(data = pdb_6zsl, format = "pdb") %>%
  m_add_label(
    text = "Label",
    sel = m_vector3(-6.89, 0.75, 0.35),
    style = m_style_label(
      backgroundColor = "#666666",
      backgroundOpacity = 0.9
    )
  ) %>%
  m_add_label(
    text = "Label2",
    sel = m_vector3(6.89, -0.75, -0.35),
    style = m_style_label(
      backgroundColor = "#000000",
      backgroundOpacity = 0.9
    )
  ) %>%
  m_zoom_to()

The positioning seems to work fine when given a residue number, but I'd like to be able to place labels outside of the molecular structure and connect them to the relevant location with an arrow. Also, m_vector3 works fine in other situations.

Thank you for the great package, it's really nice to use and makes protein vis in R easy!

I was wondering if maybe you know a way to color structures by b-factor using r3dmol? I've been checking 3dmol but I don't think they have implemented that functionality yet.
Unfortunately, I don't know any javascript and I am guessing this would have to be used as a colorfunction in the m_style_cartoon() function.

Do you maybe have any suggestions for this?

Hi,

Thanks for making r3dmol. It's a really nice package and easy to use.

I'm using m_grid in a Shiny app with selectInput. But the output graphic is not updated when changing the input in the selection menu on the left.

The individual graphics are updated, but not the multi graphic.

And I get this error on the console:
Warning in renderWidget(instance) :
Ignoring explicitly provided widget ID "c6ce055b66d1daf14e50"; Shiny doesn't use them

Any idea why? and a possible solution?

Thanks in advance :)

Example below:

library(shiny)
library(r3dmol)

ui <- fluidPage(
  sidebarLayout(
    sidebarPanel(
      selectInput("selected_id", "Choose a PDB ID:", choices=c("3G5X","7A2A","5IJ5")),
    ),
    mainPanel(
      ## Choose multi or individual, but not both at the same time
      r3dmolOutput(outputId = "structure_multi"),
      # r3dmolOutput(outputId = "structure_individual"),
      )
    )
  )
  
server <- function(input, output, session) {
  
  output$structure_individual <- renderR3dmol({
    r3dmol() %>%
      m_add_model(data = m_fetch_pdb(input$selected_id)) %>%
      m_zoom_to() %>%
      m_set_style(style = m_style_cartoon(color = "spectrum"))
    
  })
  
  output$structure_multi <- renderR3dmol({
    m1 <- r3dmol() %>%
      m_add_model(data = m_fetch_pdb(input$selected_id)) %>%
      m_zoom_to()
    
    m2 <- m1 %>%
      m_set_style(style = m_style_cartoon(color = "spectrum"))
    
    m3 <- m1 %>%
      m_set_style(style = m_style_stick())
    
    m4 <- m1 %>%
      m_set_style(style = m_style_sphere())
    m_grid(
      viewer = list(m1, m2, m3, m4),
      control_all = TRUE,
      viewer_config = m_viewer_spec(
        backgroundColor = "black"
      )
    )
    })
}


shinyApp(ui, server)

Find out how to pass JS function from R side to JS.

Enable multiple-viewer formatting like that demonstrated in the 3Dmol.js docs.

Again I tried having a go but don't know enough about javascript to have any success. Any advice would be appreciated.

Hi,

Great package! I would like to color structures with customized values, like discussed in #45. The colorfunc argument is only available in m_style_cartoon but not other representation styles. I wonder if it is possible to add this argument to m_style_surface(), m_style_sphere() and m_style_stick().

Thank you!

Please briefly describe your problem and what output you expect.

I would expect CRAN to be reasonably up-to-date.

m_button() can't work well in rmarkdown post in v0.1.2.
Buttons are layouted to the edges bounded by the whole post, not the edges bounded by widgets.

p <- r3dmol(id = "demo_sdf", elementId = "demo_sdf") %>%
m_add_model(data = benz, format = "sdf") %>%
m_set_style(style = m_style_stick()) %>%
m_set_style(sel = m_sel(model = 0),
style = m_style_stick(colorScheme = "cyanCarbon")) %>%
m_zoom_to()

output$pic <- renderPlot({(p)}, res = 96)

I am using this R package for a new shiny app construction in my paper

I find when i render this picture, it only shows in the Rstudio Plot window instead of the web

How can i fix that problem?

Thank you so much !

Ps
this was a great and very useful R package for me ! Thanks again for this amazing work !

Hi swsoyee!

Firstly, great package. Once I got the hang of it, I have been really enjoying being able to visualise structures and analysis that I've been doing in 3D in my rmkarkdown documents.

Secondly, I don't really know any javascript, only R, so excuse any glaring errors that might end up breaking the package.

I found myself always having to look up the specifics of the documentation when trying to select atoms or add styles.

I've written a number of functions, that simply return a list, but give options for the user so they know what they can specify without having to lookup the documentation of 3dmol.js. I also made some that allowed for importing from bio3d objects and fetching directly from the PDB using bio3d as an intermediate. I know 3dmol.js has that option when using url calls, but I couldn't see a way set up yet using r3dmol.

example as such:

r3dmol() %>% 
  m_add_model(data = m_fetch_pdb("1bna")) %>% 
  m_set_style(style = list(cartoon = m_cartoon_style(color = 'spectrum'), 
                           stick = m_stick_style(), 
                           sphere = m_sphere_style(scale = 0.3))) %>% 
  m_zoom_to()

I've detailed them in this vignette and they're all uploaded on my own fork of this repo.

What do you think about the functions, to allow ease of use? I've found them very useful in my own work since starting to make them.

Do you think something like this might be implemented into the base r3dmol?

@nnguyen19 Thanks for your feature request.
I have tried to implement the mapAtomProperties function, but I can't get any differences visually between with and without using mapAtomProperties.

Since I don't have an academic background in this area, I don't really understand what this function is used for.

Could you provide an example of what the visual difference is when using and not using this function?
So I can use it to test if my code is implement correctly.

When rendering a widget with {xaringan} slides, the widget resizes strangely.

In the example below, when rendered, each time the browser window is resized, the widget resizes as well. The problem arises because the container the widget canvas resizes, then the container that the canvas is is also resizes, leading to exaggerated shrinking and growing of the widget.

---
title: "Protein Example"
output: 
  xaringan::moon_reader:
    css: xaringan-themer.css
    seal: false
    nature:

      highlightStyle: github
      highlightLines: true
      ratio: 16:9
      countIncrementalSlides: true
      navigation:
        scroll: false

---

## Protein Viewer

```{r echo=FALSE, message=FALSE}
library(r3dmol)
model <- r3dmol(width = 400, height = 400) %>%
  m_add_model(r3dmol::pdb_1j72) %>%
  m_set_style(m_style_cartoon(color = "spectrum")) %>%
  m_zoom_to()
model
```

Widget after render.

Widget after shrinking the browser window.

Widget after enlarging the browser window.

I have tested it with {plotly} and the same problem doesn't happen - so it is something specific with r3dmol / 3dmol.js that is causing the problem. I just don't know enough javascript to track it down.

> sessionInfo()
R version 4.1.0 (2021-05-18)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur 11.4

Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.1/Resources/lib/libRlapack.dylib

locale:
[1] en_AU.UTF-8/en_AU.UTF-8/en_AU.UTF-8/C/en_AU.UTF-8/en_AU.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods  
[7] base     

other attached packages:
[1] ggplot2_3.3.5 r3dmol_0.1.2 

loaded via a namespace (and not attached):
 [1] pillar_1.6.2         compiler_4.1.0      
 [3] sysfonts_0.8.5       tools_4.1.0         
 [5] digest_0.6.27        jsonlite_1.7.2      
 [7] evaluate_0.14        lifecycle_1.0.0     
 [9] tibble_3.1.3         gtable_0.3.0        
[11] viridisLite_0.4.0    pkgconfig_2.0.3     
[13] rlang_0.4.11         rstudioapi_0.13     
[15] cli_3.0.1            DBI_1.1.1           
[17] yaml_2.2.1           xfun_0.25           
[19] withr_2.4.2          showtextdb_3.0      
[21] xaringanthemer_0.4.0 knitr_1.33          
[23] dplyr_1.0.7          httr_1.4.2          
[25] generics_0.1.0       vctrs_0.3.8         
[27] htmlwidgets_1.5.3    DT_0.18             
[29] grid_4.1.0           tidyselect_1.1.1    
[31] glue_1.4.2           data.table_1.14.0   
[33] R6_2.5.1             fansi_0.5.0         
[35] plotly_4.9.4.1       xaringan_0.22       
[37] rmarkdown_2.10       purrr_0.3.4         
[39] tidyr_1.1.3          magrittr_2.0.1      
[41] scales_1.1.1         ellipsis_0.3.2      
[43] htmltools_0.5.1.1    rsconnect_0.8.24    
[45] showtext_0.9-4       assertthat_0.2.1    
[47] colorspace_2.0-2     utf8_1.2.2          
[49] lazyeval_0.2.2       munsell_0.5.0       
[51] crayon_1.4.1      

Unsure how it would be implemented, but enabling the creation of buttons through functions that would resemble like that below (on the docs page:

I'm not wanting the complete functionality of {r3dmol} exposed in button form, but maybe enabling the creation of a button that could switch between cartoon or stick representation would good.

I am unsure if it is possible to implement this or not. I had an attempt and was unable to get it working, so your expertise would be appreciated.

I don't have much sense about LOGO design, do you have any idea about this topic? @BradyAJohnston

Hi, @BradyAJohnston

Sorry about bothering you again.
The m_png looks like failed to export as the figure as plot.

The output

The plot

rm(list=ls()); gc()
library(bio3d)
library(r3dmol)

p <- r3dmol() %>%
  m_add_model(data = m_fetch_pdb("6D7U", save.pdb = TRUE)) %>%
  m_add_model(data = m_bio3d(pdb)) %>%
#   m_set_style(style = m_style_cartoon(color = 'spectrum', ribbon = TRUE)) %>%
  m_set_style(style = m_style_cartoon()) %>% 
  m_add_outline() %>% 
#   m_set_style(style = c(m_style_cartoon(color = "spectrum", ribbon = TRUE), m_style_line())) %>% 
  m_add_surface(atomsel = m_sel(chain =c('A', 'B')), style = m_style_surface(opacity = 0.9, colorScheme="chain")) %>% 
  m_add_style(sel = m_sel(resi = c(206:220), chain="E"), style = m_style_sphere(scale = 0.5, color = "#e41a1c")) %>%
#   m_set_style(sel = m_sel(resi = c(206:220), chain = "A", byres = TRUE), style = m_style_cartoon(color = "red")) %>%
  m_add_surface(atomsel = m_sel(resi = c(206:220), chain = "C"), style = m_style_surface(opacity = 0.9, color = "#e41a1c")) %>%
#   m_add_line(start = m_sel(resi = 1:10, chain = "A"), end = m_sel(resi = 1:10, chain = "B"), spec = list(dashed = TRUE)) %>%
  m_add_label(text = "Site A", sel = m_sel( resi = 206:220, chain = "C"), style = m_style_label(borderColor = "green", borderThickness = 1, inFront = FALSE)) %>%
#   m_zoom_to(sel = m_sel(resi = c(206:220), chain = "A"))
#   m_set_zoom_limits(upper = 200, lower = 30)
    m_zoom_to() %>%
## Convert widgets to PNG image
   m_png()

p

R version 4.0.5 (2021-03-31)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19042)

Matrix products: default

locale:
[1] LC_COLLATE=Chinese (Simplified)_China.936 
[2] LC_CTYPE=Chinese (Simplified)_China.936   
[3] LC_MONETARY=Chinese (Simplified)_China.936
[4] LC_NUMERIC=C                              
[5] LC_TIME=Chinese (Simplified)_China.936    

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] forcats_0.5.1   stringr_1.4.0   dplyr_1.0.6     purrr_0.3.4    
 [5] readr_1.4.0     tidyr_1.1.3     tibble_3.1.1    ggplot2_3.3.3  
 [9] tidyverse_1.3.1 r3dmol_0.1.2    bio3d_2.4-1    

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.6        lubridate_1.7.10  ps_1.6.0          assertthat_0.2.1 
 [5] digest_0.6.27     utf8_1.2.1        mime_0.10         IRdisplay_1.0    
 [9] R6_2.5.0          cellranger_1.1.0  repr_1.1.3        backports_1.2.1  
[13] reprex_2.0.0      evaluate_0.14     httr_1.4.2        pillar_1.6.0     
[17] rlang_0.4.11      uuid_0.1-4        readxl_1.3.1      rstudioapi_0.13  
[21] htmlwidgets_1.5.3 munsell_0.5.0     shiny_1.6.0       broom_0.7.6      
[25] compiler_4.0.5    httpuv_1.6.1      modelr_0.1.8      pkgconfig_2.0.3  
[29] base64enc_0.1-3   htmltools_0.5.1.1 tidyselect_1.1.1  fansi_0.4.2      
[33] withr_2.4.2       crayon_1.4.1      dbplyr_2.1.1      later_1.2.0      
[37] grid_4.0.5        jsonlite_1.7.2    xtable_1.8-4      gtable_0.3.0     
[41] lifecycle_1.0.0   DBI_1.1.1         magrittr_2.0.1    scales_1.1.1     
[45] cli_2.5.0         stringi_1.6.1     fs_1.5.0          promises_1.2.0.1 
[49] xml2_1.3.2        ellipsis_0.3.2    generics_0.1.0    vctrs_0.3.8      
[53] IRkernel_1.2      tools_4.0.5       glue_1.4.2        hms_1.0.0        
[57] parallel_4.0.5    fastmap_1.1.0     yaml_2.2.1        colorspace_2.0-1 
[61] rvest_1.0.0       pbdZMQ_0.3-5      haven_2.4.1  

Yang

When do we want to push the next official release to CRAN? What sort of roadmap? Should it b1 0.1.2 or 0.2? We've added a lot of great features and I'd love to get it out to CRAN for people to use 😄