this entire repo could be a markdown file github.com/DietrichGebe...

GitHub - DietrichGebert/ponytail: Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.

Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote. - DietrichGebert/ponytail

github.com

I found where they deinterleave the trains

A bit intrigued how useful the combination of a bit of coloring, formatting + a really fast terminal app (alacritty), can be for visually inspecting reads. This one had me realize these are all pretty much the same (contaminating) sequence 👇

Taxonomy Time Machine now shows merges taxonomy.onecodex.com

the astronauts are on the dark side quick everyone hide

💾 Prokka 1.15.6 is released! This is the last major release of Prokka. But don't be sad, because @oschwengers.bsky.social already has an excellent replacement called Bakta you can migrate to. #bioinformatics #microbiology #genomics github.com/tseemann/pro...

🚨New preprint out! We present a foundational genomic resource of human gut microbiome viruses. It delivers high-quality, deeply curated data spanning taxonomy, predicted hosts, structures, and functions, providing a reference for gut virome research. (1/8) www.biorxiv.org/content/10.1...

When you buy a cutting board from bioinformaticians

gut fauna

Tech snow day

Apple's approach to protein structure is great for accessibility - & potentially biological realism - reasons. Eg, prediction could be achieved w/ smaller compute & the generative nature of prediction allows for multiple conformations A summary here: genomely.substack.com/p/simplefold...

SimpleFold and the Future of Protein Folding

A Generative Shift in Protein Folding

genomely.substack.com

If you're wondering why we're hosting the pre-print via dropbox, its because arXiv (and bioRxiv) did not accept it (because it is a review). Its a bit disconcerting, because a review is precisely the type of paper that would benefit a lot from pre-publication dissemination and feedback.

Thank you folks for your feedback on our survey about Hash functions in genomic sequence analysis. We've updated the paper and you can see the new version here: tinyurl.com/4kk9ccmt.

Closed my eyes for a sec and summoned another earthquake

they should invent a type of volatile memory that gets heavier the more data it contains

Blogged about how zstd --long fills the gap between fast and slow-but-high-ratio genome compression methods log.bede.im/2025/09/12/z...

Bede Constantinides@bede.im · 11mo ago

Zstandard's --long range mode works wonders for assemblies, but needs uninterrupted single line sequences. *AllTheBacteria 661k, multiline fasta* gzip (pigz): 751GB zstandard --long: 641GB (30% original size) *Single line fasta* gzip (pigz): 700GB zstandard --long: 232GB (10% original size)

NCBI BLAST can now output a proper CSV with headers 👏🎉: www.ncbi.nlm.nih.gov/books/NBK131...

BLAST+ Release Notes

Note: If building BLAST+ from source code, the Zlib, Zstandard and Bzip2 libraries will be needed.

ncbi.nlm.nih.gov

you can just pour milk over trail mix and eat it like cereal

Little writeup on the speed of fasta parsers, at last. Basically: both needletail and paraseq are process input linearly, and thus have a limit around 4 GB/s. By giving each thread its own slice of the input file, we're limited by RAM bandwidth instead :) curiouscoding.nl/posts/fasta-...

I do not enjoy that we now live in a world where seeing this banner at the top of PubMed makes me nervous.

TIL the EBV genome is *included in the hg38 assembly* so that EBV reads are not erroneously mapped elsewhere to the human genome. That's certainly .... an interesting solution ... 🤯 But it enabled this extremely cool work:

Caleb Lareau@caleblareau.bsky.social · last yr.

In a cool twist of fate, the EBV contig is in hg38 to mop up unscrupulous EBV reads, a by-product of immortalizing lymphoblastic cell lines (used in 1000 Genomes Project, etc.). Hence, a simple `samtools view` could get a measure of persistent EBV DNA in large WGS cohorts, e.g., UK Biobank. 4/

This is a bad take stevensalzberg.substack.com/p/i-know-gen... Saying that DNA data is like your browsing data and can can therefore be leaked is a false equivalence. Thing A is on fire so it's fine for thing B to be on fire, too-style argumentation.

I know genomes. Don't delete your DNA

Too many people are panicking about 23andMe.

stevensalzberg.substack.com

Q: what do viruses and potatoes have in common? A: both are "acellular root"

Handy to keep up with the ICTV's changes to virus taxonomy and species names: taxonomy.onecodex.com/taxon/694009... vs: taxonomy.onecodex.com/taxon/694009... or taxonomy.onecodex.com/taxon/11676/... vs taxonomy.onecodex.com/taxon/11676/...

Austin Richardson@agdr.org · 2y ago

NCBI's Taxonomy changes over time. We built Taxonomy Time Machine to track these changes: 🕰️ app: taxonomy.onecodex.com 📄 pre-print: www.biorxiv.org/content/10.1...

Are you attending #ASMicrobe this is week? Stop by my talk on Friday morning (10AM) and say hello! 👋 if you can’t make it and want to meet up - just drop me a DM! I love this meeting and connecting with so many friends and colleagues over the years has made it really a special meeting.

🌳 Taxonomy Time Machine now supports batch lookups! Quickly resolve lists of names/TaxIDs to their current NCBI taxonomy → taxonomy.onecodex.com/bulk-resolver

🚀 Pushed some updates to taxonomy.onecodex.com - Example queries to help you get started - Summary section for easier interpretation - Perf. improvements

🧵 The ATCC Genome Portal hit 5,500 authenticated microbial genomes (>2,600 species)! 🎉🥳 We've sequenced, assembled, annotated 4,538 bacteria, 479 viruses, 479 fungi, and 4 protists! All NGS in-house @ ATCC under ISO, and >90% on BOTH @nanoporetech.com and #Illumina 😎 www.atcc.org/applications...

Something happened to my $PATH and now nothing works Trisolarans: “the Sophons have succeeded in disrupting science”

Bad day for VCF files

It's clearly a DNS issue, but overall, the NCBI is the least reliable I've ever experienced in my career. And I'm in this long enough to remember the Entrez API giving you only part of the file every 50-100th time.