Oh man, there's an official Wallace and Gromit 24/7 stream on the tube.
https://www.youtube.com/live/uuxbXceM7_U?si=_VVALYy4-Zz9ie3C
Remote
naturepoker
@naturepoker@genomic.social
mastodon 4.7.3Amateur biologist at Binomica Labs. Studying Archaea from the Pangean era. Sequencing and assembling odds and ends.
#amateurbiology #microbiology #evolution #phylogeny #phylogenetics #phage #sequencing
#archaea #bioinformatics
931 Followers
1236 Following
35 Posts
Joined November 02, 2022
My lab note:
Publications:
Home page:
Our research group:
Replying to
@gvwilson@mastodon.social I feel similarly. The 'no LLM at all costs' voices in the media are leaving a big wide gap that could have used positions calling for restraint and cutting down on hype.
How the current generation of llms were developed and are being used is revolting, but the comments you've gotten here so far don't look a ton better.
Now that I have a better understanding of my own #Julialang usage patterns, time to improve my workflow.
Precompiling certain environments, writing some new functions for dotfiles and integrating everything to helix-tmux setup. I'm liking what I see so far! Will need to do an extensive writeup soon.
Good grief, looking at even HDD prices these days hurts my soul.
Picked up a used copy of The Penguin Book of the Prose Poem while coming back from work. So happy it didn't get drenched in the sudden rainstorm.
I've been really enjoying Rilke's Dinggedicht pieces recently - curious to see if there's a prose poem match for the practice.
I've also been noticing more younger audience with poetry books lately - is this a new trend? If so it's a welcome one
It looks like I'll have to withdraw my poster for ASM2026. The academic partner who's supposed to provide the PacBio reads apparently haven't even processed 1 sample in the last YEAR. 365 days.
I intentionally held off on releasing my ONT based analysis (finished November of 2025) so they can be part of the paper.
Some people from #academia seem to think collaboration with a non-academic counterpart is an act of charity, not real science where you need to pull your own damn weight.
Lol I think vs code is planning to insert copilot as co-author on git commits by default, regardless of AI use.
MS up to their old shenanigans, I see.
https://github.com/microsoft/vscode/pull/310226
This is why I shied away from vs code for #julialang and other stuff. FOSS editors and development environments all the way
I'm going with helix for LSP modal editor choice in my bioinformatics-Julialang stack since it gets me most of the way to an okay vim/emacs setup without the learning curve.
I do see myself leaning into emacs someday, but today I have work & deadlines to worry about :-/
Holy molly, the midnight weather outside is hotter than some of the mid summer days we've had before. Thursday's going to be dangerously hot from what I'm seeing. Be careful out there folks. #nyc
I gotta tell ya fellas. Working up a motivation to finish sequencing when the lab is 97F+ is getting a bit difficult.
Replying to
@adriano@lile.cl @gvwilson@mastodon.social ah yes. How stupid and short sighted of me. I'm terribly sorry to expose you, an insightful and superior being, to my lesser take. If only I had the benefit of your wisdom!
Calling #bioinformatics folks.
I'll be getting a lower end NAS like ugreen dxp4800 to offload some storage burden for upcoming ONT sequencing project (100+ pod5+analysis).
I'm curious about specific setups people have for working with NAS in #linux first environments. How it's physically setup for local network, distro choice and quirks, and etc etc. Looking for personal experiences as well as links to any write-ups out there.
Any advice would be appreciated!
일제강점기때 동립운동중 독일로 망명해 그곳에서도 반 나치 인사들과 교류하고 협력한 문학가 이미륵의 '압록강은 흐른다'가 (원작명 Der Yalu fließt) 절판에 copyright 때문에 영문 인터넷에선 거의 사장 당한듯 하다. 이게 누구를 위한 법이고, 그에 의한 새 세대의 우민화는 누가 책임질 것인가?
인종차별에 얼룩진 해외에서 수학문제나 좀 풀어 똑똑하다는 칭찬이나 듣는것이 인생의 명령이된 우리 불쌍한 이민자의 아이들은 해외에서도 문학가요 철학자들이던 이런 분들의 경험을 더 확실히 접할 필요가 있다.
Replying to
@thatdnaguy@genomic.social right!? Maybe it's a temporary thing? But it's on what looks like the official Wallace & Gromit channel. 24/7 streaming.
I have it running in the background at the lab now 😄
Speaking of - anyone know if there's any documentation/literature on how ONT put together their bacterial dorado polishing model?
I know it fares well in benchmarks, but curious about non-bacterial prokaryotes. Is anyone putting together their own polishing models for test?
Ah, but a Korean genre fiction even making the Grand Prix de l’Imaginaire short list is a great achievement I think.
The bird that drinks tears is a masterpiece of the particular type of Korean novel. There should be an English translation this summer too, I'm really looking forward to seeing how they handled the language specific word plays and world building.
Man, this is genuinely depressing.
I had to work outside shifts hauling cement bags on painkiller to pay for the travel and lodging for ASM. The days I'm there I'm losing pay and getting side eyes at work - there's no summer vacation for people like me.
Yet I managed to do everything and more. And the academic partner sends a message two weeks before the conference as if I work for them - on my own study no less! They're merely contributing data in a field they have no experience in.
It looks like our beloved @genomic.social instance is in a bit of a dry spell.
If you're a regular user and have the means, let's try to get a bit of donation heading @Psy_Fer_@genomic.social 's way! Alternative is being at the mercy of increasingly shady corpo infrastructure out there o_o
@Psy_Fer_@genomic.social jeez, I'm sorry, I went through a pretty dry spell last year (a university lab tried to claim 'experience is pay' on an actual job with contract) so went null on the donation for a long time.
Maybe I can turn some of regular binomica labs income to (very) partially supporting this instance regularly. I really want to see this instance go on for a long time to come!
4AM, last minute uploading bunch of new assembled archaeal genomes to genbank. Binomica labs represent, woot.
Replying to
@axbom true - not to mention, what regulation? AI companies had been taking whatever they want whenever they want for the last 6 or so years.
Replying to
@josephguhlin@sci.kiwi ooh I'll definitely be catching this on YouTube. Looking forward to it!
Replying to
@lwpembleton@genomic.social I also have people to see there, so I'm a bit hung up on what to do, alas.
Replying to
By the time it gets to layer 18 (the output from which gets fed into direct signal-to-sequence translation CTC), the two models end up with a very different looking output.
One thing I'm looking forward to playing with - the new #ubuntu workshop tool.
I'm curious about writing up bunch of tomls that can replicate my work & dev setup across different machines without minimal fuss. I have a feeling this might make things easier when working with other group/people's machines or running more experimental stuff that might mess with the underlying system.
Replying to
@lwpembleton@genomic.social maybe, it's just going to be a different poster from the initial abstract missing the other half of data. I'm looking into some sort of emergency sequencing, crossing fingers there's something!
Replying to
@Karhutar@mastodontti.fi @thatdnaguy@genomic.social I doubt any of this is AI (if so I'll be devastated). Wallace and Gromit is a classic claymation series, hand sculpted clay figures taken shot by shot IRL and compiled into animation
https://en.wikipedia.org/wiki/Wallace_%26_Gromit
@nwcs@mastodon.social not really, goal is to keep all read-writes to the main workstation nvme, only finished & organized research products go on the NAS.
Man I have to say it. 2026 is the worst World Cup I've seen in my life as far as Korea's concerned.
#Archaea & #microbiology folks - I'm curious. How do you feel about the dorado polish bacteria mode for your archaeal ONT long read assembled genomes? I know it holds up well with bacteria, but archaea's always been a bit funny.
Polypolish-pypolca combo is great, but they need short reads and I want to stick with long reads whenever I can (purely for logistic reasons)
Replying to
Main difference being that weights trained on complete genome reference keeps on looking toward other distant signals to make a call on current input, while contig level genome reference trained model constantly focuses on immediate signal neighbors more often than not.