I'm not real and I deny I won't heal unless I cry.let it gomore quotes

# circles: beautiful

In Silico Flurries: Computing a world of snow. Scientific American. 23 December 2017

# visualization + design

Cover image for the human genetics special issue. Trends in Genetics October 2012, 28 (10) (lowres, hires, Trends in Genetics)

# Creating the Trends in Genetics October 2012 Cover

Lately, I've been making a lot of square things round. So when Rhiannon Macrae, the Editor of Trends in Genetics, requested a Circos-like cover image for the human genetics special edition of the journal, I started drawing circles.

The image was published on the cover of Trends in Genetics human genetics special issue (Trends in Genetics October 2012, 28 (10)).

## Tools

Circos, Circos tableviewer, Illustrator CS5, and a cup (or two) of Galileo coffee from a Rancilio Epoca.

## Other Covers

Circos has appeared on covers of journals and books. Some of the images were designed by me and others were drawn from papers published in the issue.

### Journals

Cover of Blood, 2 Aug 2012, 120(5). Figure from Egan, J. B., C. X. Shi, et al. (2012). Whole-genome sequencing of multiple myeloma from diagnosis to plasma cell leukemia reveals genomic initiating events, evolution, and clonal tides. Blood 120(5): 1060-1066. (Blood)
Genomics, Aug 2012, 100(2). Figure from Katapadi, V. K., M. Nambiar, et al. (2012). Potential G-quadruplex formation at breakpoint regions of chromosomal translocations in cancer may explain their fragility. Genomics 100(2): 72-80. (Genomics)
Science Translational Medicine, December 2010, 2(61). Figure from Lo, Y. M., K. C. Chan, et al. (2010). Maternal plasma DNA sequencing reveals the genome-wide genetic and mutational profile of the fetus. Sci Transl Med 2(61): 61ra91 (Science Translational Medicine)

EMBO Journal, May 2009, 28(9). Cover design by Martin Krzywinski. (EMBO)
Nature Biotechnology, November 2009, 27(11). Figure from Cho, B. K., K. Zengler, et al. (2009). The transcription unit architecture of the Escherichia coli genome. Nat Biotechnol 27(11): 1043-1049. (Nature Biotechnology)
Genome Research, April 2008, 18(4). Cover design by Ryan Morin (Genome Research)

American Scientist, September/October 2007. Cover design by Martin Krzywinski — how it was done. (American Scientist)

### Books

iGenetics, 3rd ed. by Peter Russell (Benjamin Cummings). Cover design by Martin Krzywinski. (iGenetics)
Building Bioinformatics Solutions with Perl, R and MySQL (Oxford University Press). Cover design by Martin Krzywinski. (Building Bioinformatics Solutions)
Designing Universal Knowledge by Gerlinde Schuller (Lars Müller Publishers) (Designing Universal Knowledge)

Chromosomes — art book of film stills, David Cronenberg. Contribution to book design by Martin Krzywinski. (Chromosomes)

## source of design

I have a collection of unpublished Circos posters and thought these might be a good starting point. Rhiannon and I narrowed the choice down to the black-and-white design that showed sequenced organisms. We also liked the complex style of a panel of hundreds of Circos images generated with the tableviewer.

An old Circos poster. (zoom)
A panel of images generated from the Circos tableviewer. (zoom)

The idea would be that the foreground would be more artistic and stylized, while the background was more technical and complex. I have thousands of images available from the tableviewer (e.g. huge 15,129 image matrix).

Rhiannon also wanted to include the quote by Henry David Thoreau, "Nature and human life are as various as our several constitutions. Who shall say what prospect life offers to another?" This reminded me of a similar but more tragic line from Shakespeare's Julius Caesar, "How many ages hence shall this our lofty scene be acted over in states unborn and accents yet unknown!"

## early comps

In the early comps we played around with the idea of using non-genomics elements in the image, such as coins. We thought that we could use the variety of color and shape of the coins to communicate the idea of genetic diversity. However, after wrestling with how to do this effectively the concept was scrapped — the idea of using coins felt both arcane and arbitrary.

First set of comps. (zoom)

I decided to go with a warm brown color scheme. It's not a color I use a lot of, which makes me think that I should try to do more with it.

Deep brown provides great contrast for saturated colors, though I had to be careful not to make the image look too kitchy with an excess of colour variation. In some of the early comps shown above, two or more different color palettes were used (e.g. grey/red/blue and false color) and this lowered to overall visual cohesion of the image.

It's always a good idea to add variety to design. After all, without any variety we'd be left with a blank page. Ok, so variety is good, but too much variety is very bad, and can make you wish for that blank page again. Think about this: one kind of variety already provides variety! A variety of variety (I run the risk of recursing myself ad infinitum) can not only compete for attention but resonate destructively (that's design-speak for "turn into visual mush").

## refining the design

Everyone liked the combination of bright colors and dark background. This is an approach I favour too, which has worked well on other covers.

Experimenting with an organic look. (zoom)

Briefly I experimented with various brush and pencil filters to give the image a more hand-drawn and organic look. Most of the illustrations I generate are very digital — blocks of solid colors and high-contrast shapes — and I thought a departure from this look could work in this case. However, like with the coins, this path didn't produce anything productive.

Refining color palettes. (zoom)

## final image elements

The background is created from a matrix of about 1,400 individual Circos images created by the user community using the tableviewer. (zoom)
The main element is a Circos image of a 15 x 15 table, also created with the tableviewer. (zoom)

A watermark made up from elements in a tableviewer image that show aggregate statistics for each row and column. (zoom)
A multi-crop zoom of the main element shown above. This version is colored for punch. (zoom)

Masks showing the locations of smaller vignettes. (zoom)
An 8 x 8 tableviewer image with outlined ribbons. (zoom)

Thoreau quote: Nature and human life are as various as our several constitutions. Who shall say what prospect life offers to another? (zoom)

Background and midground elements. (zoom)
Background and foreground elements. (zoom)

## final image

Final image with all the layers. (Trends in Genetics October 2012, 28 (10)) (zoom)
VIEW ALL

# Statistics vs Machine Learning

Tue 03-04-2018
We conclude our series on Machine Learning with a comparison of two approaches: classical statistical inference and machine learning. The boundary between them is subject to debate, but important generalizations can be made.

Inference creates a mathematical model of the datageneration process to formalize understanding or test a hypothesis about how the system behaves. Prediction aims at forecasting unobserved outcomes or future behavior. Typically we want to do both and know how biological processes work and what will happen next. Inference and ML are complementary in pointing us to biologically meaningful conclusions.

Nature Methods Points of Significance column: Statistics vs machine learning. (read)

Statistics asks us to choose a model that incorporates our knowledge of the system, and ML requires us to choose a predictive algorithm by relying on its empirical capabilities. Justification for an inference model typically rests on whether we feel it adequately captures the essence of the system. The choice of pattern-learning algorithms often depends on measures of past performance in similar scenarios.

Bzdok, D., Krzywinski, M. & Altman, N. (2018) Points of Significance: Statistics vs machine learning. Nature Methods 15:233–234.

Bzdok, D., Krzywinski, M. & Altman, N. (2017) Points of Significance: Machine learning: a primer. Nature Methods 14:1119–1120.

Bzdok, D., Krzywinski, M. & Altman, N. (2017) Points of Significance: Machine learning: supervised methods. Nature Methods 15:5–6.

# Happy 2018 $\pi$ Day—Boonies, burbs and boutiques of $\pi$

Wed 14-03-2018

Celebrate $\pi$ Day (March 14th) and go to brand new places. Together with Jake Lever, this year we shrink the world and play with road maps.

Streets are seamlessly streets from across the world. Finally, a halva shop on the same block!

A great 10 km run loop between Istanbul, Copenhagen, San Francisco and Dublin. Stop off for halva, smørrebrød, espresso and a Guinness on the way. (details)

Intriguing and personal patterns of urban development for each city appear in the Boonies, Burbs and Boutiques series.

In the Boonies, Burbs and Boutiques of $\pi$ we draw progressively denser patches using the digit sequence 159 to inform density. (details)

No color—just lines. Lines from Marrakesh, Prague, Istanbul, Nice and other destinations for the mind and the heart.

Roads from cities rearranged according to the digits of $\pi$. (details)

The art is featured in the Pi City on the Scientific American SA Visual blog.

Check out art from previous years: 2013 $\pi$ Day and 2014 $\pi$ Day, 2015 $\pi$ Day, 2016 $\pi$ Day and 2017 $\pi$ Day.

# Machine learning: supervised methods (SVM & kNN)

Thu 18-01-2018
Supervised learning algorithms extract general principles from observed examples guided by a specific prediction objective.

We examine two very common supervised machine learning methods: linear support vector machines (SVM) and k-nearest neighbors (kNN).

SVM is often less computationally demanding than kNN and is easier to interpret, but it can identify only a limited set of patterns. On the other hand, kNN can find very complex patterns, but its output is more challenging to interpret.

Nature Methods Points of Significance column: Machine learning: supervised methods (SVM & kNN). (read)

We illustrate SVM using a data set in which points fall into two categories, which are separated in SVM by a straight line "margin". SVM can be tuned using a parameter that influences the width and location of the margin, permitting points to fall within the margin or on the wrong side of the margin. We then show how kNN relaxes explicit boundary definitions, such as the straight line in SVM, and how kNN too can be tuned to create more robust classification.

Bzdok, D., Krzywinski, M. & Altman, N. (2018) Points of Significance: Machine learning: a primer. Nature Methods 15:5–6.

Bzdok, D., Krzywinski, M. & Altman, N. (2017) Points of Significance: Machine learning: a primer. Nature Methods 14:1119–1120.

# Human Versus Machine

Tue 16-01-2018
Balancing subjective design with objective optimization.

In a Nature graphics blog article, I present my process behind designing the stark black-and-white Nature 10 cover.

Nature 10, 18 December 2017

# Machine learning: a primer

Thu 18-01-2018
Machine learning extracts patterns from data without explicit instructions.

In this primer, we focus on essential ML principles— a modeling strategy to let the data speak for themselves, to the extent possible.

The benefits of ML arise from its use of a large number of tuning parameters or weights, which control the algorithm’s complexity and are estimated from the data using numerical optimization. Often ML algorithms are motivated by heuristics such as models of interacting neurons or natural evolution—even if the underlying mechanism of the biological system being studied is substantially different. The utility of ML algorithms is typically assessed empirically by how well extracted patterns generalize to new observations.

Nature Methods Points of Significance column: Machine learning: a primer. (read)

We present a data scenario in which we fit to a model with 5 predictors using polynomials and show what to expect from ML when noise and sample size vary. We also demonstrate the consequences of excluding an important predictor or including a spurious one.

Bzdok, D., Krzywinski, M. & Altman, N. (2017) Points of Significance: Machine learning: a primer. Nature Methods 14:1119–1120.