PCA-app, tab #3: Fit

[Lars Skjærven]

A PyMol session download or VMD state file would be awesome here (same for PCA and NMA tabs)!

Coloring options should include 'variability' (RMSF), and 'alignment position' (index).

The heatmap display can suffer when there are lots of structures and having no interactivity or sense of values here is a real shame. See: http://kbroman.org/qtlcharts/example/iplotCorr.html We can do this with some of the JS libs and this would also be cool for other heatmap occurrences.

[Lars Skjærven]

I'll do the PyMOL sessions download on this issue

[Barry Grant]

I can do VMD one (once I see how your PyMol one looks). I will also do the webGL coloring by variability and alignment position.

[Lars Skjærven]

pymol session download added to fit tab. at the moment coloring options are: cluster id, structure id, core, and gap regions.

todo: import alignment to pymol session.

[Lars Skjærven]

BUG in fit tab for > 300 structures: For type=1: Color vector should be the same length as the number of structures

[Barry Grant]

Thanks for catching this Lars. I will add it to the list for view.xyz()

[Lars Skjærven]

iplotCorr looks good, but I have trouble including it to shiny, and I can not separate it from the scatterplot on the right hand side.

# server
output$qtl_rmsd_heatmap <- iplotCorr_render({
  pdbs <- fit()
  rd <- rmsd1()
  rownames(rd) <- pdbs$lab
  colnames(rd) <- pdbs$lab
  hc <- hclust(as.dist(rd))
  grps <- cutree(hc, k=input$clusters)
  iplotCorr(rd)
})

# ui 
iplotCorr_output("qtl_rmsd_heatmap")