You don’t have to be an APBRmetrician to appreciate the graphics Kirk Goldsberry puts together from his analysis. They’re clear, the data-to-ink ratio is high, and there are appropriate annotations to help tell the story.
Previously on Labrigger: Visualizations
“The Journal Impact Factor, as calculated by Thomson Reuters, was originally created as a tool to help librarians identify journals to purchase, not as a measure of the scientific quality of research in an article.”
Signed by HHMI, Wellcome, AAAS, EMBO, eLife, PLoS, …
PubPeer (highlighted previously on Labrigger) caught some shenanigans that apparently the reviewers and editors at Cell did not. Here’s the comment from PubPeer, and here’s the paper.
Figure 2F (top image) and the left side of Figure 6D (second image) are the same set of cells, but are presented as different in the paper.
Here’s a second example of image/data reuse from the same paper. Figure S6 reused the same exact data for two panels. The two panels even appear right next to each other in the figure.
I find WPI’s Kwik-Cast handy for some things around the lab. When we were modest users, the price didn’t bother me too much. However, when we started going through so much of it, I wondered if there was a way to buy it that fit our consumption level a bit better.
Both of those pots together were about $50. (Body Double Fast Set – Trial Size, Reynolds Advanced Materials)
These are the double-barreled syringes and mixing tips I bought. They’re from this company.
2mm x 8 Element, Needle Tip
qty. 1-99 – $0.91 each (WPI charges about $2.90 per tip, about 318% more)
qty. 100+ – $0.637 each (Buy 100 for $63.70. From WPI, 30 tips cost $87.00, 455% more)
4B19 Double barreled syringe
qty. 1-99 – $1.953 each
qty. 100+ – $1.367 each
Here, you can see below that the WPI Kwik-Cast kit uses the same syringes and mixing tips, at least as far as I can tell.
From WPI, each of these syringes (with contents) is $85.
So you can buy the DIY stuff above ($50), 50 tips ($9.10 * 5 = $45.50), and 10 syringes ($19.53) for a grand total of $115.03. This would cost $850 + ($29 * 5 = $145) = $995.00 from WPI.
Plus, with the DIY way, you’ll still have the vast majority of your silicone elastomer left over from the $50 kit. I’m not sure I’ll ever run out.
The drawback of the DIY way is that you have to fill your own syringes, or have an undergrad do it for you. I used a couple of 5 mL syringes, and it took maybe a minute or two to do one syringe.
Undocumented MATLAB has an in depth look at the next generation graphics handler for MATLAB which you can use today, although it’s not officially released yet. Use the command line option “-hgVersion 2″ when launching MATLAB. See the post for more details.
Here’s a follow up on the previous post about alternatives to Google Reader (which is being shut down).
Patrick Mineault commented that Feedly is looking good. I agree. Basic functionality is smooth and somewhat intuitive; layout and design are excellent.
Post by Jeffrey Stirman
The opacity of the brain is one barrier to optically imaging individual neurons and their connections. Scattering in tissue is the main reason tissue is not transparent; absorption also plays a role but much less so. Perfusing tissue with a substance to match the index of refraction throughout the preparation (and thus decrease scattering) is one approach, and although index matching isn’t a new strategy, just getting rid of the membranes is. The most recent method to achieving tissue transparency (Chung et al., 2013), takes this approach to great effect.
A nice paper discussing tissue transparency is Johnsen and Widder, 1999. Scattering in tissue is dominated by Mie scattering which is the scattering of light by particles of a size on the same order as the wavelength of light (Rayleigh scattering is for particles much smaller than the wavelength): cells, nuclei, and organelles all fit in this category. Furthermore, the lipid membranes encasing these structures have a significantly different refractive index (~1.5) than the surrounding medium. It is this change in refractive index of these particles that lead to scattering. Simply, as the difference in refractive index between the surrounding medium and the object increases, so too does the scattering. The relationship with wavelength can be complicated and range from about lambda^-4 to lambda^0.2 (lambda = the wavelength of light used) depending on the size of the particle, but overall the higher the wavelength, the less scattering (one of the benefits of 2-photon imaging).
A couple of nice papers from Mourant et al. (1998 & 2000) discuss and explore in more detail the dominant scattering centers in tissue. They found that at small angles, most of the scattering was dominated by the nucleus and at larger angles the smaller structures such as mitochondria. One conclusion from all this is perhaps it might not be sufficient to homogenize the refractive index of the tissue if those lipid membranes still exist (as earlier attempts had done). In fact, the best way to achieve tissue clarity for imaging is to remove the objects that cause the scattering. This is exactly what Kwanghun Chung did! By first crosslinking most of the proteins, DNA, and other biological entities (not the lipids), then cross-linking them all in a hydrogel structure, he was able to use a detergent extraction process (electric field assisted) to remove the lipid membranes and thereby removing the cause of most of the scattering centers. Since multiple rounds of antibody staining can be performed on the cleared tissue, this process seems to have achieved clarity while preserving most of the interesting biology.
