Kasper D Hansen

It’s the time of the year where 98% of my feed will become posts from @michaelhoffman.bsky.social on ASHG

Illustration of Burrows-Wheeler Transform and many auxiliary structures from the input string how$now$brown$cow$#

New tool "bwt-svg" for making illustrations of the BWT and the many auxiliary arrays and other structures related to it. Pyodide-based no-installation-necessary interface here: benlangmead.github.io/bwt-svg/. (H/t to @robert.bio for pointing me to pyodide!) Full repo: github.com/benlangmead/....

Identification of de novo variants from parent-proband duos via long-read sequencing While de novo variants cause many Mendelian disorders, their detection currently requires sequencing of the proband and both biological parents. This is not feasible when only one parent is available,...

Looking forward to presenting duoNovo at ASHG this week.

We use long-read sequencing to detect de novo variants without having to sequence both parents. It's conceptually straightforward, and performs very accurately among variants likely to be clinically relevant.

www.medrxiv.org/content/10.1...

Screenshot one of two of blog post at https://dirk.eddelbuettel.com/blog/2025/10/10#rcpparmadillo_15_transition_office_hours detailing offer of office hours to help with CRAN deadlines for packages depending on RcppArmadillo

Screenshot two of two of blog post at https://dirk.eddelbuettel.com/blog/2025/10/10#rcpparmadillo_15_transition_office_hours detailing offer of office hours to help with CRAN deadlines for packages depending on RcppArmadillo

RcppArmadillo 15 CRAN Transition: Offering Office Hours

Offering office hours to help in updating #rstats CRAN packages to RcppArmadillo 15.0.2 given CRAN 'Deadlines' for a number of the reverse-dependency packages. Details and booking link at blog post.

dirk.eddelbuettel.com/blog/2025/10...

Amazing opportunity for help if you're using RcppArmadillo in a CRAN package.

Thanks for posting. This was an amazing read.

This is an important limitation and something we should absolutely work on solving.

I enjoyed reading the Introduction.

Despite the fact that the cover features neither a dragon, nor a spaceship, I have ordered a copy.

Books And Their Covers - Penny Arcade Videogaming-related online strip by Mike Krahulik and Jerry Holkins. Includes news and commentary.

How can we know if it is worth reading without seeing the cover? I hope it has either a dragon or a space ship.

www.penny-arcade.com/comic/2022/0...

This is a masterpiece. Simply excellent
www.instagram.com/reel/DMVGMIC...

Didn’t want to believe this was true. It’s true.

Higher eQTL power reveals signals that boost GWAS colocalization Expression quantitative trait locus (eQTL) studies in human cohorts typically detect at least one regulatory signal per gene, and have been proposed as a way to explain mechanisms of genetic liability...

Good discussion & study-design recommendations in this new preprint by @mikelove.bsky.social @klmohlke.bsky.social & colleagues:
Higher eQTL power reveals signals that boost GWAS colocalization 🧪🧬

www.biorxiv.org/content/10.1...

It all depends on how many times a piece of code is run. Yes, for highly user workflows, it makes a lot of sense to optimize. But that is not all code.

My first conference ever was at IPAM as an undergrad, bringing me into scientific research.

When I branched out into new topics, my first conference as new faculty was also at IPAM.

My R35 was suspended. A lot of personal feelings, but more I worry for the future of the country.

“There’s no better example of how [HIV and COVID were] different than to look at what happened to Tony Fauci. Celebrated for leading the response to HIV, pilloried for playing the same leading role with COVID, he eventually needed a security service to protect him and his family”

Nick Eagles on analyzing Visium HD at #GBCC2025

Stefano @stemang.bsky.social showing off #tidyomics at the #GBCC2025 !

Yeah this is a (really) stupid idea. It is a core feature of science that we cannot do this. Whoever proposed this does not understand research.

Quote of the day from #GBCC2025
“It can fail.
But, sometimes, it works”

Most DNN get run for a fixed number of iterations. There is no clear concept of convergence.

Smyth uses log2( (x+0.5) / (L+1) ) if I recall correctly.

I will be there.

Will this be settled unarmed or with rapiers? I prefer to travel light unless it’s necessary :)

Upon reflecting, I think it is fair that the biggest issue with Slack threads is how notifications are handled. And that is an implementation / design issue. Threads in Slack seems to be meant as a side conversation between a few specific users.

Worth reading

I just want one channel with no threads

I fundamentally disagree that a channel is defined by people. In my view, which I do think is how slack is designed (but I could be wrong), a channel is defined by a topic.

And if a channel needs extensive threading, the topic is too broad.

I've seen folks talk about how "in four years" they'll be able to get back to their planned projects and I really don't think that's going to work out. People & projects losing funding now will not be able to hit pause and come back once funding is restored. Hard-won progress & capacity will be lost

I recognize that my gut reaction is likely wrong, but I don't really see the point of threads. If you need threads, shouldn't you just make a new channel? It is not clear to me how this scales over time where you may have 100s or 1000s of threads. Do they expire?

I am at StatGen 2025 in Minneapolis. I would be happy to chat with students or postdocs about research or career questions. Send me a DM.

Very interesting post.