Scripts to create cartoons of 3D genomes on blender.

• R-base
• RANN R package
• blender v2.77. Have not tested other versions but it might work unless something really basic had changed.
• Does not require but makes use of the color palettes from colorbrewer.

https://vimeo.com/genome3d

Based on the concepts developed on "Capturing pairwise and multi-way chromosomal conformations using chromosomal walks".

In general one has to create the semi-random walk that will determine the "chromatin". This is done on R. After saving the chromatin path in R one should import it using the python script within blender. A handful of blender files are provided and links to the corresponding videos too.

source('tads.r')
# for a rather dense large "TAD"
gg = generate_tad_random(6000, warp = 200, limit = 0.7, abs_limit = 0.9 * 4)
# save the path as a text file
write.table(gg$path, "path_6k_dense_w_200_l_0.7.txt", quote=F, row.names=F, sep="\t");

Once the path is computed it has to be imported into blender. The code at this moment has a lot of room for improvement but for the time being one must manually change the path at the bottom of the chromosomes.py script and point towards the text file generated in R.

tad_path6 = importWalk("path_6k_dense_w_200_l_0.7.txt")

The current scripts by no means represent real information. They are just a cartoon of how I imagine things to work. This of course could be fed with real data (not sure yet what type) and draw something more "real".

keywords: 3d genome, hic, chromatin, chromosome conformation capture, 3C, chromosomal walks, C-walks.