@ab604.uk
Developing micromixers for Hydrogen Deuterium Exchange Mass Spectrometry Proteomics. Occasional immunopeptidomics researcher at the University of Southampton. https://ab604.uk/ #rstats 🤖 My proteomics literature bot: @protpapers.bsky.social
Great article. Not sure whether R/tidyverse is your preferred language, but I've curated some R skills for Claude Code with rules and commands to help with efficient token usage and enforce constraints that may be of interest.
Based on others work, I've created Claude Code configurations for R: modular skills (tidyverse, rlang, performance, OOP, testing), enforcement rules (security, testing, git workflow), workflow commands (planning, code review, TDD), and context management hooks. #claudecode #rstats
Hi Jeremy. I really appreciated what Sarah and yourself had done. It inspired me to create a Claude Code repo of my own incorporating eight R skills, commands and rules e.g. for security, hooks for token use efficiency and agents e.g. code review. I'll keep updating. github.com/ab604/claude...
I went for a long walk yesterday on the Isle of Wight. I took so many photos I created a post about it.
Currently reading Once was Willem by MR Carey. Not finished but good so far. A Drop of Corruption by Robert Jackson Bennett is the best published in 2025 fiction book I've read, but it's the 2nd of the series so start with Tainted Cup. They're genre novels though.
They're super helpful, thank you.
Trying to write a new section for our (stalled) lung neoantigens paper and finding these two articles by @dingdingpeng.the100.ci very helpful. The marginal effects opening paragraph always makes me laugh.
www.the100.ci/2024/12/01/w...
www.the100.ci/2022/05/27/%...
#30DayChartChallenge Day 6 Florence Nightingale theme. A hexagonal density plot, where each hexagon show the vaccination coverage for local authorities in England for 5 year olds in 2024 and the colour indicates the coverage percentage. Where reported, the the lowest coverage was 75% and the highest 100%. One acute measles-related death was reported in 2024 in a young person who was known to have other medical conditions.
#30DayChartChallenge Day 6: Florence Nightingale.
A hexagonal density plot England of measles vaccination coverage for 5 year olds. There was one measles related death in England in 2024.
Vaccines work.
I know it's a rough time to be talking science, but in case anyone wants some distraction, I wrote about the Tan et al (2024) analysis of heritability and polygenic indices for complex human behavior. Title: Is Tan et al The End of Social Science Genomics?
Scatterplot of April 2025 US tariffs for 196 countries plotted against a guess of how the tariff was calculated using the trade balance for each country divided by their import amount in dollars divided by 2. For example, China has a balance with the USA of 295 billon and the US imports 439 billion worth of goods, which gives a guess tariff of 34% which is the same as the actual tariff. All EU countries have 20% tarriff imposed. This re-creates an interactive chart by FT Alphaville: https://on.ft.com/4cddkuC and the code and data I used for each country are at: https://github.com/ab604/charts-2025
For #30DayChartChallenge Big or Small, I couldn't resist re-creating the FT Alphaville chart on US tariffs: on.ft.com/4cddkuC. See the article for their interactive version.
All the tariffs are big, but some are bigger than others!
Code and country data is in my repo: github.com/ab604/charts...
Image of the moon phase on the 3rd of April 2025 in Southampton in the UK where it is a waxing crescent moon, 19% of the way through it's cycle. It shows a pale yellow crescent on a black background. The image is from a webpage that updates each day with the current moon phase: https://ab604.github.io/moon-phase/
For #30DayChartChallenge Circles, I chose the Moon.
I made a webpage that updates each day with an image of the moon phase as it appears from Southampton in the UK.
Today it is a waxing crescent moon, 19% of the way through it's cycle.
The wepage is at ab604.github.io/moon-phase/
Line chart for four cohorts of women screened for cervical cancer from 1989 to 2019. It reproduces an Our World in Data chart that shows how in England, younger cohorts vaccinated at school against the Human papillomavirus show dramatically lower rates of cervical cancer than older cohorts who were not vaccinated, when they reached the same age. Vaccination programmes for girls began in 2008 in England and then 2019 for boys. The DOI for the source data article used by Our World in Data is 10.1016/S0140-6736(21)02178-4
For slopes #30DayChartChallenge I've made a simpler version of the Our World in Data graphic about cervical cancer and the Human papillomavirus: ourworldindata.org/hpv-vaccinat...
Good news, vaccines work and prevent cancer!
Source article DOI is 10.1016/S0140-6736(21)02178-4
Stacked bar plots showing that England accounts for the majority of lung cancer diagnoses across all age groups in the UK and how in most age groups, people diagnosed with lung cancer are usually smokers. The data is for 45,563 patients diagnosed in the UK from January 2000 to December 2021. Source data from DOI: 10.21037/tlcr-24-241, Made by: Alistair Bailey, Code: https://github.com/ab604/charts-2025
Trying the #30DayChartChallenge for the first time. I'm going to focus on my research interests, so here's plots of UK lung cancer diagnoses between 2000 and 2021 showing how most people who get lung cancer are smokers, and how England has the largest proportion of cases.
Not sure about proud, but I did write about how I made my Quarto website ab604.uk/how
Overdiagnosis of mental disorders is making headlines again, this time thanks to Wes Streeting. But it is true?
(thread)
www.theguardian.com/politics/202...
Lastly, I've not read it, but heard good things and may discuss the philosophy is Tom Chivers recent book: www.amazon.com/Everything-P...
And if you're more interested in discussion then www.the100.ci is awesome
And also www.bayesrulesbook.com and bayesf22.classes.andrewheiss.com are also great for Bayes.
With Richard McElreath's Statistical Rethinking and his lectures would be one great place to start: www.youtube.com/watch?v=FdnM...
Realised I didn't actually link to the study on fonts and reading speed directly. It's here: dl.acm.org/doi/10.1145/...
I did a brief summary of how "there is no best font" last year: ab604.github.io/webpage-desi... and if you want to go straight to the study it's here: www.nngroup.com/articles/bes... Currently I favour Inter.
New post!
"To never risk a bad holiday or a crap meal is to experience a kind of living death. Without risk there is no learning, no growth, no surprise, no delight, no adventure, no anecdote for the dinner table, just the bland homogenous tyranny of ‘good’."
Sea foam on the beach at Southbourne at sunset. The foam is blurred due from long exposure.
And I like the light on the sea at this time of day.
Photo of the sunset from the beach at Southbourne looking towards Bournemouth. Red sky fading to blue with the sea and sand.
Beautiful sunset at Bournemouth this evening.
Hope this is not too off topic, but I thought this post by @rachelhart.co.uk was excellent on the subject of language and deserves a wide audience: rachelhart.substack.com/p/sectioned-...
TIL that there is a legal alternative to SciHub for accessing academic articles: unpaywall.org . It's a browser extension for Firefox or Chrome and I wrote a short post: ab604.uk/blog/2025-02...
New post just out:
"Technology vs Democracy"
On the origins of Muskworld's ideology, which it's so much more dangerous than merely stupidity or corruption, and what might stop it.
(£/free trial)
open.substack.com/pub/samf/p/t...
Not sure if this is relevant to your problem, but I bookmarked this some time ago for when I need reminding why one plots raw p-values: support.bioconductor.org/p/98442/
And this was the comparison paper I came to it by, but I think there is a MCP paper for the package too: www.nature.com/articles/s41...