Combining Art and Science

This is a confession. I love cells. However, if this were an essay for my German teacher at secondary school she would, given the name of this blog, award me a “Themenverfehlung” (basically completely missing the point of the essay question), which she did on a semi-regular basis. I could argue that in the broadest sense this post is also about cells, because cells were necessary for the creation of this work of art as well as for the analysis. But I admit that’s a bit of a stretch.

Anyway. I found this new paper published in Scientific Reports (Thurrowgood et al, 2016), which is simply entitled, ‘A Hidden Portrait by Edgar Degas’. The Australian research group, comprised of art curators and X-ray technology specialists, used a technique called X-ray fluorescence to study “Portrait of a Woman”, which hangs in the National Gallery of Victoria, Australia. Normal X-ray analysis had already revealed that an earlier portrait of a different woman lies underneath what is currently visible:

figure 1

Visible image of “Portrait of a Woman” (left) and conventional X-ray imaging (right) – copied from Figure 1

Conventional X-ray technology relies on the different densities of the materials/pigments in the painting. The denser areas absorb more of the X-rays and the X-ray film behind is less exposed than behind less dense areas, just like in a chest X-ray, for example. However, X-ray fluorescence is a more sensitive technique because it can detect specific elements in the painting. When atoms are hit by high energy X-rays some of the electrons close to the atomic nucleus (the “inner shells”) can be flung out of the atom entirely. This leaves a hole for electrons to fill and when these electrons “fall” into the inner shells energy is released in the form of electromagnetic waves, which are characteristic for different atoms:


Schematic of X-ray fluorescence – copied from this link

This technique is not entirely new, but as far as I understand this was the first time it has been used for a large-scale analysis of a painting. Using the physical properties of different atoms, Thurrowgood et al. created so-called elemental maps to determine which parts of the hidden portrait contained which elements.

figure 3

Elemental maps – copied from Figure 3

By analysing which elements are found in the same areas the researchers could trace back which pigments were used and therefore make educated guesses about the original colour of the painting. For example, the hair contained a lot of iron and manganese, which are often found together in the brown umber pigment.

figure 6

False colour image reconstruction of the hidden portrait of Mlle. Dobigny – copied from Figure 6

Mademoiselle Dobigny was one of Degas’ favourite models, but who knows why he kept her portrait for so long before deciding to paint over it. The reasons are certainly less obvious than in Georges Seurat’s painting “Young Woman Powdering Herself” (Courtauld Institute of Art, London): he concealed the only known self-portrait of his behind a mirror, watching the young woman who also happened to be his lover…

Reference and further reading:

Thurrowgood D, Paterson D, de Jonge MD, Kirkham R, Thurrowgood S, Howard DL (2016) A Hidden Portrait by Edgar Degas. Scientific Reports 6: 29594

New York Times article on Degas’ painting

Courtauld Institute of Art blog on Seurat’s painting

The #IceBucketChallenge Two Years On

August two years ago saw our facebook feeds flooded with footage of friends and acquaintances dousing themselves in ice-cold water to raise awareness and money for amyotrophic lateral sclerosis (ALS; or motor neuron disease (MND)) charities. It was the topic of the inaugural blog post and a follow-up one year later. My inherently slightly cynical and skeptical nature questioned whether all this social media craze (and £87.7 million raised) would actually make a difference. Well, facebook came to the rescue and linked me to articles from the BBC and The Guardian alerting me to a paper published recently in Nature Genetics (Kenna et al., 2016).

The researchers contributing to this study work in eleven countries across the world. (Who ever thought science benefited from international collaboration? Am I still frustrated by Brexit? No, not at all…) Large proportions of the funding were provided by the National Institute of Health in the USA as well as  ALS and MND Associations in the USA and UK. Kenna et al. sequenced those parts of the genome that are actively expressed – a technique known as whole exome sequencing – in over 1000 familial/inherited ALS patients and over 7000 controls. Since sequencing technologies are becoming better and cheaper all the time, this is the less impressive part of the study. Next, all this data was processed using so-called gene burden analyses. This is where I stop understanding what is done with the data, but in essence it was possible to use previously known genetic risk factors of ALS to infer overlooked genes that are also associated with the disease. In the figure below the genes indicated in blue are genes that were already known to confer ALS risk (e.g. SOD1 and FUS), whereas those in black are the new genes, and everything above the red dotted line was considered statistically significant.


Graph depicting genes associated with ALS risk – copied directly from Kenna et al., 2016

As you can see, the researchers identified mutations in the NEK1 gene through these sequencing and data analysis experiments. However, only about 10% of people with ALS have the familial/inherited form of the disease. Therefore Kenna et al. then went on to check whether these NEK1 mutations could also be found in samples from patients with sporadic ALS and indeed they could. Overall approximately 3% of ALS patients have abnormal NEK1 genes.

After all this data analysis the paper ends with a description of what the NEK1 protein normally does and what it might not be doing in ALS patients’ cells: NEK1 helps to repair damaged DNA and contributes to the formation of an organelle called the cilium. Now future experiments will have to focus on exactly why and how mutations in NEK1 contribute to ALS. And since only 3% of ALS patients have NEK1 mutations there are still many other genes to discover.

The Project MinE aims to do just that – with headquarters in the Netherlands, this collaborative DNA sequencing project is analysing samples from even more ALS patients and controls. Their website says that a donation of €75 enables sequencing and analysis of a single chromosome! Anyone fancy another cold shower?

Lastly, I’ve found an interesting article combining two of my favourite things (science and Impressionist art) – so look forward to that in the next post.


Kenna KP, van Doormaal PTC, Dekker AM, Ticozzi N, Kenna BJ, Diekstra FP, van Rheenen W, van Eijk KR, Jones AR, Keagle P, Shatunov A, Sproviero W, Smith BN, van Es MA, Topp SD, Kenna A, Miller JW, Fallini C, Tiloca C, McLaughlin RL et al. (2016) NEK1 variants confer susceptibility to amyotrophic lateral sclerosis. Nature Genetics advance online publication.