Interactive data visualization/plotting with D3.js and/or plotly, on and off the web

[danfulop]

I'm interested in interactive plotting, and D3 or plotly (which is partly based on D3) seem like the way to go.

plotly would be easier to use for the "narrow" task of making beautiful interactive graphs that can be published on the web. It has APIs for R, Python, and many other languages: https://plot.ly/api/ ...as well as plot converters for ggplot2 (R) and matplotlib (Python).

D3 is a much more general tool (it's a Javascript library) for "for manipulating documents based on data" and visualizing said data. For those of us that don't know javascript, it may be easier and faster to use plotly. However, eventually learning D3 would be great! ...and probably very marketable as a skill.

I for one would be very interested in visualizing RNAseq/DGE data in a map framework, as I'll soon have gene expression data from several genotypes of ~17 species distributed across South America.

Any thoughts? ...takers?

[iamciera]

I thought I had responded to this already!

I am really down for learning D3. I will help with your RNAseq mapping. I want to do something similar with my LCM RNAseq but on regions of the leaf. Kinda like what @DanChitwood already did here: http://danchitwood.github.io/CaulerpaGeneExpression/. Give it a second it takes a bit to load. The code is here:https://github.com/DanChitwood/CaulerpaGeneExpression. I imagine there are some nice D3 map setups already though for your project!

[kaisakajala]

I would love to learn how to create visualizations like this but it's totally above my head and skill level... where should I start now to eventually to be able to learn D3?

[danfulop]

@kaisakajala a near-term compromise could be using plotly's ggplot2 converter. That is, plot w/ ggplot2 in R, and then convert it to plotly. plotly is based on D3.

[iamciera]

I think D3 is over our head too! D3 is a javascript library. None of us know javascript, i'm sure you can learn along with us.

[kaisakajala]

:) I would love to at least attempt to! My data (one day) will be fairly similar to your LCM data - multiple cell types - but also a couple of different species, different abiotic stresses and multiple levels of regulation. It would be neat to have a way to visualize expression patterns in an interactive and clear way! I like Dan's clusters with Caulerpa and of course the eFP viewer, but I'd like to consider if there would be a more powerful way of displaying the data.

[DanChitwood]

Let me know what you guys find out with respect to interactive,
multivariate data visualization techniques, especially if compatible with
R/ggplot2. I'd like to add: I did have some success with d3.js. But, I
didn't know javascript, and it took a lot of time to use it in a kind of
template-based way from Mike Bostock's amazing examples. It took a lot of
time, and if you look under the hood of the few times I got things to work,
it's probably the least elegant code you've seen. So yes, d3.js can be done
for a big, specific, grand project, but it's not good for everyday life
like ggplot2. For me, ggplot2 maximizes exploration and understanding of
data for yourself. Also, if you're into interactivity, Karl Broman's
qtlcharts (http://kbroman.org/qtlcharts/) should also be on your list.

On Wed, Oct 15, 2014 at 7:07 PM, kaisakajala [email protected]
wrote:

:) I would love to at least attempt to! My data (one day) will be fairly
similar to your LCM data - multiple cell types - but also a couple of
different species, different abiotic stresses and multiple levels of
regulation. It would be neat to have a way to visualize expression patterns
in an interactive and clear way! I like Dan's clusters with Caulerpa and of
course the eFP viewer, but I'd like to consider if there would be a more
powerful way of displaying the data.

—
Reply to this email directly or view it on GitHub
#4 (comment)
.

Dan Chitwood, Assistant Member
Donald Danforth Plant Science Center
975 North Warson Road
Saint Louis, MO 63132

Room: 1064 (1st Floor, West)
p: (314) 587-1609
f: (314) 587-1709
e: [email protected]

Website: chitwoodlab.org

[iamciera]

This project subject matter has been combined with other into DjangoSiteforD3Vis repo

[iamciera]

@DanChitwood! I am totally adding you to the Coding Collective. You could help us understand what you did. We want to make a website to guide people in making a D3 visualizations of their RNAseq data. The repository is here: https://github.com/TheCodingCollective/DjangoSiteforD3Vis. If you ever wanted to get more experience collaborating on an open source project through github, this is the time.

[DanChitwood]

Thanks so much! There are certainly people here who are eager to follow,
learn, and contribute. I'll spread the word!

On Wed, Oct 15, 2014 at 7:55 PM, Ciera Martinez [email protected]
wrote:

@DanChitwood https://github.com/DanChitwood! I am totally adding you to Coding
Collective https://github.com/TheCodingCollective/Welcome. You could
help us understand what you did. We want to make a website to guide people
in making a D3 visualizations of their RNAseq data. The repository is here:
https://github.com/TheCodingCollective/DjangoSiteforD3Vis. If you ever
wanted to get more experience collaborating on an open source project
through github, this is the time.

—
Reply to this email directly or view it on GitHub
#4 (comment)
.

Dan Chitwood, Assistant Member
Donald Danforth Plant Science Center
975 North Warson Road
Saint Louis, MO 63132

Room: 1064 (1st Floor, West)
p: (314) 587-1609
f: (314) 587-1709
e: [email protected]

Website: chitwoodlab.org

[danfulop]

Hey DanC, welcome!! BTW, it was awesome to hear you on SciFri talking about Stradivari and how form doesn't follow function in violins, violas, etc! I podcast SciFri semi-religiously and it was a nice surprise to suddenly hear you come on. ...sorry to be off topic.

[DanChitwood]

Thanks so much!!! The first Science Friday episode I went to watch in
preparation for the interview was your piece on orchids!

On Wed, Oct 15, 2014 at 8:08 PM, danfulop [email protected] wrote:

Hey DanC, welcome!! BTW, it was awesome to hear you on SciFri talking
about Stradivari and how form doesn't follow function in violins, violas,
etc! I podcast SciFri semi-religiously and it was a nice surprise to
suddenly hear you come on. ...sorry to be off topic.

—
Reply to this email directly or view it on GitHub
#4 (comment)
.

Dan Chitwood, Assistant Member
Donald Danforth Plant Science Center
975 North Warson Road
Saint Louis, MO 63132

Room: 1064 (1st Floor, West)
p: (314) 587-1609
f: (314) 587-1709
e: [email protected]

Website: chitwoodlab.org