And whatever I do will become forever what I've done.don't rehearsemore quotes

# space: exciting

EMBO Practical Course: Bioinformatics and Genome Analysis, 5–17 June 2017.

# visualization + design

Cover image accompanying our article on mouse vasculature development. Biology turns astrophysical. PNAS 1 May 2012; 109 (18) (zoom, PNAS)

# Creating the PNAS Cover

One of my goals in life, which I can now say has been accomplished, is to make biology look like astrophysics. Call it my love for the Torino Impact Hazard Scale.

Recently, I was given an opportunity to attend to this (admittedly vague) goal when Linda Chang from Aly Karsan's group approached me with some microscopy photos of mouse veins. I was asked to do "something" with these images for a cover submission to accompany the manuscript.

When people see my covers, sometimes they ask "How did you do that?" Ok, actually they never ask this. But being a scientist, I'm trained me to produce answers in anticipation of such questions. So, below, I show you how the image was constructed.

The image was published on the cover of PNAS (PNAS 1 May 2012; 109 (18))

## Tools

Photoshop CS5, Nik Color Efex Pro 4, Alien Skin Bokeh 2 and a cup of coffee from a Rancilio Silvia.

## source images

Below are a few of the images I had the option to work with. These are mouse embryonic blood vessels, with a carotid artery shown in the foreground with endothelial cells in green, vascular smooth muscle cells in red and the nuclei in blue.

Of course, as soon as I saw the images, I realized that there was very little that I needed to do to trigger the viewer's imagination. These photos were great!

Mouse carotid arteries. (zoom)
Mouse carotid arteries. (zoom)
Mouse carotid arteries. (zoom)
Mouse carotid arteries. (zoom)
Mouse carotid arteries. (zoom)
Mouse carotid arteries. (zoom)
Mouse carotid arteries. (zoom)
Mouse carotid arteries. (zoom)
Mouse carotid arteries. (zoom)

## memories of star trek

Immediately I thought of two episodes of Star Trek (original series): Doomsday Machine and the Immunity Syndrome, as well as of images from the Hubble Telescope.

Enterprise is about to be consumed by a horror tube: a planet killer! (The Doomsday Machine)
Enterprise heads into a giant amoeba. Who eats whom? I'll let you guess. (The Immunity Syndrome)
Orion nebula (M42) as seen by the Hubble telescope. (zoom)

I though it would be pretty easy to make the artery images look all-outer-spacey. They already looked it.

## centerpiece image

And then I saw the image below.

A particularly spectacular image of a mouse carotid artery. I'm thinking 10 on the Torino scale. (zoom)

## constructing the cover

### background

The background was created from the two images shown here. The second image was sampled three times, at different rotations.

Images used for background. (zoom)
Images used for background. (zoom)
Layer composition for background elements. (zoom)

The channel mixer was used to remove the green channel and leave only red and blue.

Background elements for PNAS cover image. (zoom)

### middle ground

The next layer was composed of what looked like ribbons of blue gas. This was created by sampling the oval shapes from the source images. Here the red channel was a great source for cloud shapes, and this was the only channel that was kept. The hue was shifted to blue and a curve adjustment was applied to increase the contrast.

First set of middle ground elements, before adjustments. (zoom)
First set of middle ground elements, after channel adjustments. (zoom)
Second set of middle ground elements. (zoom)
Layer composition for middle ground elements. (zoom)

When the foreground and middle ground elements were combined, the result was already 40 parsecs away.

Background and foreground elements for PNAS cover image. (zoom)

### foreground

The foreground was created from the spectacular comet-like image of a mouse artery. Very little had to be done to make this element look good. It already looked good.

I applied a little blur using Alien Skin's Bokeh 2 to narrow the apparent depth of field, masked out elements at the bottom of the image and removed some of the green channel. The entire blue channel was removed altogether (this gave the tail of the comet a mottled, flame-like appearance).

Foreground element, after channel adjustments. (zoom)
Layer composition for foreground element. (zoom)

### post processing

Initial composition of background, middle ground and foreground elements. (zoom)
40% localized application of Nik's Tonal Contrast (Color Efex 4 plugin) to increase structure in red channel. (zoom)
50% blend with Nik's Pro Contrast (Color Efex 4 plugin). (zoom)

And here we have the final image.

Final PNAS cover. Spacey! (zoom)
VIEW ALL

# Classification and regression trees

Fri 28-07-2017
Decision trees are a powerful but simple prediction method.

Decision trees classify data by splitting it along the predictor axes into partitions with homogeneous values of the dependent variable. Unlike logistic or linear regression, CART does not develop a prediction equation. Instead, data are predicted by a series of binary decisions based on the boundaries of the splits. Decision trees are very effective and the resulting rules are readily interpreted.

Trees can be built using different metrics that measure how well the splits divide up the data classes: Gini index, entropy or misclassification error.

Nature Methods Points of Significance column: Classification and decision trees. (read)

When the predictor variable is quantitative and not categorical, regression trees are used. Here, the data are still split but now the predictor variable is estimated by the average within the split boundaries. Tree growth can be controlled using the complexity parameter, a measure of the relative improvement of each new split.

Individual trees can be very sensitive to minor changes in the data and even better prediction can be achieved by exploiting this variability. Using ensemble methods, we can grow multiple trees from the same data.

Krzywinski, M. & Altman, N. (2017) Points of Significance: Classification and regression trees. Nature Methods 14:757–758.

Lever, J., Krzywinski, M. & Altman, N. (2016) Points of Significance: Logistic regression. Nature Methods 13:541-542.

Altman, N. & Krzywinski, M. (2015) Points of Significance: Multiple Linear Regression Nature Methods 12:1103-1104.

Lever, J., Krzywinski, M. & Altman, N. (2016) Points of Significance: Classifier evaluation. Nature Methods 13:603-604.

Lever, J., Krzywinski, M. & Altman, N. (2016) Points of Significance: Model Selection and Overfitting. Nature Methods 13:703-704.

Lever, J., Krzywinski, M. & Altman, N. (2016) Points of Significance: Regularization. Nature Methods 13:803-804.

# Personal Oncogenomics Program 5 Year Anniversary Art

Wed 26-07-2017

The artwork was created in collaboration with my colleagues at the Genome Sciences Center to celebrate the 5 year anniversary of the Personalized Oncogenomics Program (POG).

5 Years of Personalized Oncogenomics Program at Canada's Michael Smith Genome Sciences Centre. The poster shows 545 cancer cases. (left) Cases ordered chronologically by case number. (right) Cases grouped by diagnosis (tissue type) and then by similarity within group.

The Personal Oncogenomics Program (POG) is a collaborative research study including many BC Cancer Agency oncologists, pathologists and other clinicians along with Canada's Michael Smith Genome Sciences Centre with support from BC Cancer Foundation.

The aim of the program is to sequence, analyze and compare the genome of each patient's cancer—the entire DNA and RNA inside tumor cells— in order to understand what is enabling it to identify less toxic and more effective treatment options.

# Principal component analysis

Thu 06-07-2017
PCA helps you interpret your data, but it will not always find the important patterns.

Principal component analysis (PCA) simplifies the complexity in high-dimensional data by reducing its number of dimensions.

Nature Methods Points of Significance column: Principal component analysis. (read)

To retain trend and patterns in the reduced representation, PCA finds linear combinations of canonical dimensions that maximize the variance of the projection of the data.

PCA is helpful in visualizing high-dimensional data and scatter plots based on 2-dimensional PCA can reveal clusters.

Altman, N. & Krzywinski, M. (2017) Points of Significance: Principal component analysis. Nature Methods 14:641–642.

Altman, N. & Krzywinski, M. (2017) Points of Significance: Clustering. Nature Methods 14:545–546.

# $k$ index: a weightlighting and Crossfit performance measure

Wed 07-06-2017

Similar to the $h$ index in publishing, the $k$ index is a measure of fitness performance.

To achieve a $k$ index for a movement you must perform $k$ unbroken reps at $k$% 1RM.

The expected value for the $k$ index is probably somewhere in the range of $k = 26$ to $k=35$, with higher values progressively more difficult to achieve.

In my $k$ index introduction article I provide detailed explanation, rep scheme table and WOD example.