resources
visualizations
presentations
lunch break
Distractions and amusements, with a sandwich and coffee.
statistics + data
Nature Methods: Points of Significance
Access all columns for free at Statistics for Biologists Nature Collection.
A Statistics Primer and Best Practices
The Points of Significance column was launched in September 2013 as an educational resource to authors and to provide practical suggestions about best practices in statistical analysis and reporting.
This month we launch a new column "Points of Significance" devoted to statistics, a topic of profound importance for biological research, but one that often doesn’t receive the attention it deserves.
The "aura of exactitude" that often surrounds statistics is one of the main notions that the Points of Significance column will attempt to dispel, while providing useful pointers on using and evaluating statistical measures.
—Dan Evanko, Let's Give Statistics the Attention it Deserves in Biological Research
The column is co-authored with Naomi Altman (Pennsylvania State University). Paul Blainey (Broad) is a contributing co-author.
Free Access
In February 2015, Nature Methods announced that the entire Points of Significance collection will be free.
When Nature Methods launched the Points of Significance column over a year ago we were hopeful that those biologists with a limited background in statistics, or who just needed a refresher, would find it accessible and useful for helping them improve the statistical rigor of their research. We have since received comments from researchers and educators in fields ranging from biology to meteorology who say they read the column regularly and use it in their courses. Hearing that the column has had a wider impact than we anticipated has been very encouraging and we hope the column continues for quite some time.
—Dan Evanko, Points of Significance now free access
Also, in a recent post on the ofschemesandmemes blog, a new statistics collection for biologists was announced.
The pieces range from comments, to advice on very specific experimental approaches, to the entire collection of the Points of Significance columns that address basic concepts in statistics in an experimental biology context. These columns, originally published in Nature Methods thanks to Martin Krzywinski and guest editor Naomi Altman, have already proven very popular with readers and teachers. Finally, the collection presents a web tool to create box plots among other resources.
—Veronique Kiermer, Statistics for biologists—A free Nature Collection
continuity and consistency
Each column is written with continuity and consistency in mind. Our goal is to never rely on concepts that we have not previously discussed. We do not assume previous statistical knowledge—only basic math. Concepts are illustrated using practical examples that embody the ideas without extraneous complicated details. All of the figures are designed with the same approach—as simple and self-contained as possible.
news + thoughts
Yearning for the Infinite — Aleph 2
Discover Cantor's transfinite numbers through my music video for the Aleph 2 track of Max Cooper's Yearning for the Infinite (album page, event page).
I discuss the math behind the video and the system I built to create the video.
Hidden Markov Models
Everything we see hides another thing, we always want to see what is hidden by what we see.
—Rene Magritte
A Hidden Markov Model extends a Markov chain to have hidden states. Hidden states are used to model aspects of the system that cannot be directly observed and themselves form a Markov chain and each state may emit one or more observed values.
Hidden states in HMMs do not have to have meaning—they can be used to account for measurement errors, compress multi-modal observational data, or to detect unobservable events.
In this column, we extend the cell growth model from our Markov Chain column to include two hidden states: normal and sedentary.
We show how to calculate forward probabilities that can predict the most likely path through the HMM given an observed sequence.
Grewal, J., Krzywinski, M. & Altman, N. (2019) Points of significance: Hidden Markov Models. Nature Methods 16:795–796.
Background reading
Altman, N. & Krzywinski, M. (2019) Points of significance: Markov Chains. Nature Methods 16:663–664.
Hola Mundo Cover
My cover design for Hola Mundo by Hannah Fry. Published by Blackie Books.
Curious how the design was created? Read the full details.
Markov Chains
You can look back there to explain things,
but the explanation disappears.
You'll never find it there.
Things are not explained by the past.
They're explained by what happens now.
—Alan Watts
A Markov chain is a probabilistic model that is used to model how a system changes over time as a series of transitions between states. Each transition is assigned a probability that defines the chance of the system changing from one state to another.
Together with the states, these transitions probabilities define a stochastic model with the Markov property: transition probabilities only depend on the current state—the future is independent of the past if the present is known.
Once the transition probabilities are defined in matrix form, it is easy to predict the distribution of future states of the system. We cover concepts of aperiodicity, irreducibility, limiting and stationary distributions and absorption.
This column is the first part of a series and pairs particularly well with Alan Watts and Blond:ish.
Grewal, J., Krzywinski, M. & Altman, N. (2019) Points of significance: Markov Chains. Nature Methods 16:663–664.
1-bit zoomable gigapixel maps of Moon, Solar System and Sky
Places to go and nobody to see.
Exquisitely detailed maps of places on the Moon, comets and asteroids in the Solar System and stars, deep-sky objects and exoplanets in the northern and southern sky. All maps are zoomable.
Quantile regression
Quantile regression robustly estimates the typical and extreme values of a response.Quantile regression explores the effect of one or more predictors on quantiles of the response. It can answer questions such as "What is the weight of 90% of individuals of a given height?"
Unlike in traditional mean regression methods, no assumptions about the distribution of the response are required, which makes it practical, robust and amenable to skewed distributions.
Quantile regression is also very useful when extremes are interesting or when the response variance varies with the predictors.
Das, K., Krzywinski, M. & Altman, N. (2019) Points of significance: Quantile regression. Nature Methods 16:451–452.
Background reading
Altman, N. & Krzywinski, M. (2015) Points of significance: Simple linear regression. Nature Methods 12:999–1000.
Analyzing outliers: Robust methods to the rescue
Robust regression generates more reliable estimates by detecting and downweighting outliers.Outliers can degrade the fit of linear regression models when the estimation is performed using the ordinary least squares. The impact of outliers can be mitigated with methods that provide robust inference and greater reliability in the presence of anomalous values.
We discuss MM-estimation and show how it can be used to keep your fitting sane and reliable.
Greco, L., Luta, G., Krzywinski, M. & Altman, N. (2019) Points of significance: Analyzing outliers: Robust methods to the rescue. Nature Methods 16:275–276.
Background reading
Altman, N. & Krzywinski, M. (2016) Points of significance: Analyzing outliers: Influential or nuisance. Nature Methods 13:281–282.
Two-level factorial experiments
To find which experimental factors have an effect, simultaneously examine the difference between the high and low levels of each.Two-level factorial experiments, in which all combinations of multiple factor levels are used, efficiently estimate factor effects and detect interactions—desirable statistical qualities that can provide deep insight into a system.
They offer two benefits over the widely used one-factor-at-a-time (OFAT) experiments: efficiency and ability to detect interactions.
Since the number of factor combinations can quickly increase, one approach is to model only some of the factorial effects using empirically-validated assumptions of effect sparsity and effect hierarchy. Effect sparsity tells us that in factorial experiments most of the factorial terms are likely to be unimportant. Effect hierarchy tells us that low-order terms (e.g. main effects) tend to be larger than higher-order terms (e.g. two-factor or three-factor interactions).
Smucker, B., Krzywinski, M. & Altman, N. (2019) Points of significance: Two-level factorial experiments Nature Methods 16:211–212.
Background reading
Krzywinski, M. & Altman, N. (2014) Points of significance: Designing comparative experiments.. Nature Methods 11:597–598.