Martin Krzywinski / Genome Sciences Center / mkweb.bcgsc.ca Martin Krzywinski / Genome Sciences Center / mkweb.bcgsc.ca - contact me Martin Krzywinski / Genome Sciences Center / mkweb.bcgsc.ca on Twitter Martin Krzywinski / Genome Sciences Center / mkweb.bcgsc.ca - Lumondo Photography Martin Krzywinski / Genome Sciences Center / mkweb.bcgsc.ca - Pi Art Martin Krzywinski / Genome Sciences Center / mkweb.bcgsc.ca - Hilbertonians - Creatures on the Hilbert Curve
listen; there's a hell of a good universe next door: let's go.e.e. cummingsgo theremore quotes

visualization: fun



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


communication + science

Nature Methods: Points of View

Martin Krzywinski @MKrzywinski mkweb.bcgsc.ca
Points of View column in Nature Methods. (Points of View)

Guidelines for Effective Figures

Practical and concise advise on the visual presentation of data for researchers. One topic and one page at a time.

Common Challenges in Figure Design

Andreas Dahlin runs a figure making course at Uppsala University. He was kind to share with me common questions and concerns that his students have when creating figures (emphasis is mine).

I face problems for using the tools in power point to make nice illustration figures, and in addition how one can enhance the resolution of the figures to print it in a high quality mode.

In my opinion, the most difficult thing is how to draw the good-looking pictures and design the structure of slide to make it simple and substantial in content.

I find it difficult to find the right software to draw pictures.

The most difficult thing for me, when I make a figure, is to arrange the parts of the figure in a way they look nice and understandable.

I think the most difficult part is creating the concept, how to make a figure easy and fast to understand but not lacking all essential parts.

Stepping outside of my own knowledge of what the picture presents and viewing it as someone who sees it for the first time. It's easy to assume that some things are self evident and not making them clear enough in the pictures.

Figures that not are plots can also be tricky to get to look nice.

Anytime you have to draw something in paint, gimp, or other image program it requires a lot of work to make it look even slightly better than crap.

The most difficult thing (in general) is to include as much information as possible and display it in a way that is easy to understand. Figures should be intuitive for the reader, which is sometimes difficult to achieve. There might also be technical difficulties in achieving what you've visualized.

I think the most difficult part for me is to highlight the main idea I would like to express.

For me the most difficult part is making 3-D figures. Also while making figures its hard to decide on the good colors to choose for the figure.

In my opinion, the most difficult part when making a figure is don't know which software we can use and how to use.

The most difficult part for me is to start it! Because I am so meticulous and I am a painter, then it is not so easy to make decision about my figures and which one is better and so on, then finally I give up and put just one figure which of course I don't like...

I think it is difficult to put together my ideas to something that is connected and makes it easier for the viewer to understand.

It is so easy to just get an image from internet. I don’t know what is ok to do. There seems to be different rules in different communities.

To come up with a figure that does not simplify the concept too much at the same time as it does not overwhelm the viewer. To get some ideas for this is the reason why I take the course. ;-)

To me, how to make it easy to understand is the difficult part.

I think it is to save it in the correct format: Raster or vector, png or jpg or pdf... especially if I want to make some changes in the future to the figure.

I think is to choose the most appropriate figure that really help to transmit the information we want. Then, how many words can be good enough for been part of the message. At the beginning I used to use too many.

Apart from the difficulty of making the figure clear and easy to understand, the biggest problem I'm having is the captions. How long and detailed description is appropriate, so it neither steals attention from the figure nor leaves out too much important information.

I think the most difficult part is to have high resolution image once we want to save it. My experience is when finish with drawing, the file size sometimes to large for high quality image and if we downgrade it, the image becomes bad.

The most difficult part when i making a figure is the software using part, I'm not good at computer so that part is annoying for me all the time.

I think the most difficult is to find out how to condensate many ideas in one picture without making it difficult to understand.

The most difficult part is the get the image to not look too amateurish that people focus on that instead of the message.

The most difficult part when doing a figure is to let it speak for itself, i.e. to not have long caption text.

To be able to depict all the desirable results on a single figure is sometimes not that easy. It becomes more critical when a figure is to be fitted within a certain size frame. An exact placing of a figure in some text editors often comes along with difficulties.

The most difficult part when making a figure is to make it simple and still be informative.

Depends a lot on the kind of figure, but generally it is to get clarity in the design, such that the idea is conceived easily. This requires some good outline (usually an iterative process).

The most difficult part to make a figure is the need to express abstract concepts into drawings.

The compromise between include detailed information and at the same time be readable (figures in articles)

To compress all information and ideas you have in your head into short and clear message.

I feel the difficulty in choosing a right resolution of the picture and the angle that could visualize all the details. And also choosing right test/label colour, size, font. Another difficulty for me is continuation from one slide to another.

I believe that my biggest problem would be making nice flux charts. Generally the ones I draw look too crude, it does not look beautiful. I have no concern about making an image that can represent an idea, but making a beautiful image makes it more pleasing to the eyes of the people who will read my work.

It is very difficult to make the figure delicate. I am still not get used to put all the small components together to integrate the figure by the vector software, instead of drawing it out directly.

I think the most difficult part is to make the image simple but yet informative.

I find it very difficult to make an original clarity picture in a particular format after dimensioning it according to the requirement.

Some times it is difficult to limit the size (Bytes) of the picture when going for high clarity remake.

Making the figure as informative as you want while keeping it simple enough to grasp quickly.

For me, the more difficult part is to create a figure that contains or tells all the information that I want to transmit, but keeping the figure simple, clean and not overloaded.

The most difficult for me is make it easily to be understood meanwhile containing the essential information.

The most difficult thing when developing a figure is ... to remove the bloat but keep the message. (Besides the very most difficult: finding out what I want to tell.)

For me the most difficult part is to choose colors with right contrast and to make it more attractive and catchy.

VIEW ALL

news + thoughts

Ensemble methods: Bagging and random forests

Mon 16-10-2017
Many heads are better than one.

We introduce two common ensemble methods: bagging and random forests. Both of these methods repeat a statistical analysis on a bootstrap sample to improve the accuracy of the predictor. Our column shows these methods as applied to Classification and Regression Trees.

Martin Krzywinski @MKrzywinski mkweb.bcgsc.ca
Nature Methods Points of Significance column: Ensemble methods: Bagging and random forests. (read)

For example, we can sample the space of values more finely when using bagging with regression trees because each sample has potentially different boundaries at which the tree splits.

Random forests generate a large number of trees by not only generating bootstrap samples but also randomly choosing which predictor variables are considered at each split in the tree.

Krzywinski, M. & Altman, N. (2017) Points of Significance: Ensemble methods: bagging and random forests. Nature Methods 14:933–934.

Background reading

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

...more about the Points of Significance column

Classification and regression trees

Mon 16-10-2017
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.

Martin Krzywinski @MKrzywinski mkweb.bcgsc.ca
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.

Background reading

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.

...more about the Points of Significance column

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).

Martin Krzywinski @MKrzywinski mkweb.bcgsc.ca
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.

Martin Krzywinski @MKrzywinski mkweb.bcgsc.ca
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.

Background reading

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

...more about the Points of Significance column

`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.