Julien Roux

How thoughtful experimental design can empower biologists in the omics era - Nature Communications Here, the authors discuss principles of experimental design that are relevant for all biology research, along with special considerations for projects using -omics approaches, highlighting common expe...

Fantastic review on the importance of rigorous experimental design. Amen. 😊

"'To consult the statistician after an experiment is finished is often merely to ask him to conduct a post mortem examination. He can perhaps say what the experiment died of.' – Ronald A. Fisher"

@manuelkleiner.bsky.social

How thoughtful experimental design can empower biologists in the omics era - Nature Communications Here, the authors discuss principles of experimental design that are relevant for all biology research, along with special considerations for projects using -omics approaches, highlighting common expe...

Great publication from Maggie Wagner (KU) and Manuel Kleiner (NCS) on how to design microbiome (+ other Omic) projects with downstream statistics in mind!

A must read for trainees!
www.nature.com/articles/s41...

The article has also a table with more in depth resources on various aspects of experimental design for more advance design needs.

How thoughtful experimental design can empower biologists in the omics era - Nature Communications Here, the authors discuss principles of experimental design that are relevant for all biology research, along with special considerations for projects using -omics approaches, highlighting common expe...

I am really excited to see this article on basic experimental design principles published. Definitely will be mandatory reading for incoming graduate students in my lab. Maggie Wagner who led this article did an amazing job with the illustrations of basic principles www.nature.com/articles/s41...

For example many people would think that you need to double the number of animals if you plan an experiment including both sexes and want to keep statistical power... Which is actually not true

Yes, we will discuss experimental design and power issues. Even for bioinformaticians, it's important to be able to weigh in on a discussion about study design!

One challenge is to analyze the data rigorously, for example extract significant interactions between sex and other factors of interest. Another is to then interpret the results (e.g., no significance doesn't always mean there is no effect, maybe the experiment is underpowered...)

Correction: our course will now be taught online!

Our course will actually be taught online, to make it easier for people abroad to attend!

Half-half, but we'll plan backup options for the hands-on part for people with limited R knowledge!

Newcomers are more than welcome of course! A bit of background on the analysis of genomics and transcriptomics data would be useful however

Our invited speakers for the parallel session on ‘Computational methods for single-cell and spatial omics’ is @stephaniehicks.bsky.social from Johns Hopkins Bloomberg School of Public Health, USA. Session is chaired by @marianna-raps.bsky.social and @julienroux.bsky.social. #bc2basel‬‬‬‬‬‬‬

🗓️📍Date and location: 27 October 2025, Bern, Switzerland

Contact us if you need more info!
@sib.swiss

Ideally, data analysts should be involved even before data is generated, so you need to be ready to take on this role! 🧑‍🔬
During the day we'll guide you through the steps of designing robust genomics experiments, and rigorously analysing them to draw valid conclusions. 🖥️ 🧬

Best Practices for Investigating Gene Expression Differences by Biological Sex Overview Sex and gender have been historically ignored as explanatory variables in transcriptomics studies, in particular because of the fear that add

Registrations now open for our course 𝐁𝐞𝐬𝐭 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐞𝐬 𝐟𝐨𝐫 𝐈𝐧𝐯𝐞𝐬𝐭𝐢𝐠𝐚𝐭𝐢𝐧𝐠 𝐆𝐞𝐧𝐞 𝐄𝐱𝐩𝐫𝐞𝐬𝐬𝐢𝐨𝐧 𝐃𝐢𝐟𝐟𝐞𝐫𝐞𝐧𝐜𝐞𝐬 𝐛𝐲 𝐁𝐢𝐨𝐥𝐨𝐠𝐢𝐜𝐚𝐥 𝐒𝐞𝐱 www.sib.swiss/training/cou...

💡Why attend? More and more policies require attention to sex and gender in research projects, but incorporating this variable is challenging!

Welcome everyone to the 2025 [BC]² Basel Computational Biology Conference. Speakers and attendees from all over the world gather for 3 exciting days of discussion to shape the future of data-driven biology where #Bioinformatics meets #AI. #bc2basel

The figure shows how summer temperature has increased in CH since 1980 and how heat related mortality developed during that period.

Just released by @swisstph.ch: new figures of the heat-related mortality monitoring in Switzerland: In summer 2024 (6th warmest in CH), around 326 deaths were attributed to heat, less than in summers 2022 and 2023. South was mostly affected.
www.nccs.admin.ch/nccs/en/home...

I'm told that @eseb2025.bsky.social is communicating on X/Twitter. I call on the organisers to leave it now. Being there reinforces the network, we don't have to do that. Pin a message sending to Bluesky (even Mastodon? @eseb.bsky.social is there) and people will follow. #ESEB2025

European Bioconductor Conference 2025 – EuroBioC2025 The European Bioconductor Conference (EuroBioC2025) will take place on September 17th to 19th, 2025, in Barcelona, Spain. EuroBioC2025 will bring together the Bioconductor community to showcase the…

⏳ One week left to register!
Join us at EuroBioC2025 in Barcelona and the Bioconductor Carpentry Workshops for hands-on learning, community building, and cutting-edge bioinformatics.

🗓️ Registration closes August 28 — don’t miss out!
🔗 eurobioc2025.bioconductor.org

#EuroBioC2025 #RStats

Congrats! What are you going to do with all these genomes? 🤯

CrossFilt: A Cross-species Filtering Tool that Eliminates Alignment Bias in Comparative Genomics Studies Comparative functional genomic studies are often affected by biased read mapping across species due to inter-species differences in genome structure, sequence composition, and annotation quality. We d...

We developed CrossFilt, a read-level, cross-species filtering approach that avoids these errors without sacrificing usable data. It outperforms all existing strategies and exposes just how much signal in comparative genomics has been misread.

doi.org/10.1101/2025...

"IN MICE" is back! Now with bonus yelling! He actually had a noticeable effect on medical reporting with the Old Blue Site "justsaysInMice" account through encouraged ridicule.

Worth a follow if you care about basic accuracy in medical reporting 8-)

But sometimes…
<begin rant>
... You realize that the authors have no intention to share anything with the community ("data available upon reasonable request").

This slows science down so much! We simply can’t build on previous work without the data behind results shown in papers! 💩😡
</end rant>

(So I've uploaded pre-processed data to GEO — GSE273902 — and the code to reproduce all figures from there is on Github: github.com/julien-roux/...)

Sometimes human data (e.g., raw sequencing reads) cannot be shared due to patient privacy. Still, it’s super helpful to share aggregated data such as read count tables! We’re actually in this exact situation as the patients in our cohort did not sign up a consent allowing data sharing 🙃

📊 A word about the comparative analysis:
A big part of this study was to compare our results to other transcriptomics datasets, in the context of obesity, type 2 diabetes, bariatric surgery, etc.

Some datasets were a dream to work with: raw data was public, clean, well-annotated.
Others, less so…

At some point, I thought that this deserved to be out as a paper, so I decided to take it as a “side project” — on top of all routine work on the table (if there is such thing as routine in this job 😆)
Having this out now feels really rewarding! 🎉

A PI in my department brought in this dataset in 2021, originally generated by collaborators even before. I did the analysis as usual, but with noone was a clear lead on the project so things stalled...

A bit of backstory on this paper:

As a bioinformatician in a core facility, it’s rare to be first author on a paper, so this project is special!

This was an exciting dataset to work on, I learned so much about obesity and T2D. And I had a lot of bioinformatics fun along the way too!
This paper shows the power of good study design and rigorous analysis 🧬🖥️✨
Feedback welcome before we submit to some journal! 📨🤡
13/13