Alexey Sergushichev

How about letting people decide for themselves when to disconnect or when not?

I had a similar, less principled way of doing this, but this does seem to work quite well after trying a couple of large changes with it: boristane.com/blog/how-i-u...
(I use codex, but same idea)

And how do you pay for it? Lab members reimbursed for individual subscription? Team plan for the lab? Any discounts?

PhD programs worldwide face an urgent question: How should trainees use ChatGPT, Claude & similar tools?

Now online: two resources on thoughtfully integrating generative AI into research training.

📄 Conceptual framework: zenodo.org/records/18649847
📋 Practical guide: zenodo.org/records/18452319

🧵

COMBINE-lab - The skeptic’s guide to generative AI assisted coding An easy-to-use, flexible website template for labs, with automatic citations, GitHub tag imports, pre-built components, and more.

I’ve written a post about my recent experiences (successes) with AI coding models; the experiences that caused me to re-evaluate my initial judgements, the surprise I had at what can be accomplished, & some fears I have about these tools. Discussion welcome! combine-lab.github.io/blog/2026/02...

Please fix the preview screen, it says "template" :)

Postdoctoral Researcher or Senior Scientist (AI × Biology)

Boeva Lab is #hiring postdocs and senior scientists in AI/ML for Cancer Biology! Check the offer and apply at: jobs.ethz.ch/job/view/JOP...

Dealing with ScienCV formatting Some tips for composing and making your new NIH biosketch not look like garbage

Been hearing some horrifying chatter about dealing with the new #NIH common form and #ScienCV system, so I put together a short video with some tricks for mitigating some of those pain points.

Hope it helps? 🧪

open.substack.com/pub/emptymod...

Somehow I got a lot of duplicate publications in ScieENcv. Does anyone know what's best way to fix it? Does it matter at all?

I've seen programming languages in Russian (my native), and it feels very cringy. English language for programming is much smoother for some reason.

More info and examples are on the package web-site: alserglab.github.io/mascarade/ Source code is on GitHub: github.com/alserglab/ma... 3/3

For this release I optimized the performance and added helper functions for working with Seurat objects 2/3

mascarade R package for generating borders on UMAP/t-SNE plots is out on CRAN! 🧬🖥️ 1/3

Start parsing from all shifts in parallel, but withhold the data until you end up with one shift? Probably shouldn't be a big overhead if done only in the beginning of a large chunk

if you're interested, note that the time is UK time. In the US, that will be 9:00am.

I'm just delighted to announce our new preprint on genome-scale perturb-seq in CD4+ T cells. We learned both general lessons about the power of perturb-seq, and specific lessons about T cell biology.

Led by amazing postdocs Emma Dann and Ronghui Zhu, with my wonderful collaborator Alex Marson.

In my view, the repo is the most actual state of the package and that the package built from there should pass the check, so it is very related. Unlike the working directory, which contains an intermediate state.

But it seems like a non-mainstream view based on the discussion here and on github.

But this goes with using untracked files and other "fundamental issues". While I'm more interested in checking that CI/CD won't fail because of some stupid mistake I made. And since rcmdcheck doesn't mirror the CI/CD behavior, running it isn't as helpful for that.

Am I missing something? AFAIU, rcmdcheck doesn't know anything about git and uses the local dir. I have no problem of building a temp tar.gz to check. My custom script does exactly this: copies git tree, applies staged diff, builds tar.gz, and checks it. This mirrors CI/CD behavior, but locally.

Am I the only #rstats package developer annoyed that I can't easily run devtools::check() on the current version of the git tree (+staged)? Hate that options are either pushing it to CI/CD, or running a custom script. Anyone to support the corresponding devtools issue github.com/r-lib/devtoo...?

Causal modelling of gene effects from regulators to programs to traits - Nature Approaches combining genetic association and Perturb-seq data that link genetic variants to functional programs to traits are described.

GWAS has been an incredible discovery tool for human genetics: it regularly identifies *causal* links from 1000s of SNPs to any given trait. But mechanistic interpretation is usually difficult.

Our latest work on causal models for this is out yesterday:
www.nature.com/articles/s41...
A short🧵:

Looks like AI slop

Role play in question: "You are an experienced software engineer. Write a program that ..."

CATS: Canine Airport Therapy Squad

Bioinformatics is in shambles at the Denver airport’s acronym game

Technically correct, the best kind of correct

I find the lack of reference to Star Wars with this name disturbing

Motherfucking Rex in the motherfucking park

📢 We are announcing the Adaptive Immune Profiling Challenge 2025!
Can you predict immune state labels from adaptive immune receptor repertoires?
💰 $10,000 prize pool!
🗓️ Launches Nov 5 on @kaggle.com
More Info: uio-bmi.github.io/adaptive_imm...

Does anyone have a high-res version of NCBI GEO logo (www.ncbi.nlm.nih.gov/geo/img/geo_...)? I hate putting their low-res version in the slides/figures. I've tried to use AI tools for upscaling, but the results aren't perfect (colors are off). Anyone want to take a shot at it? Here is my best take: