Moritz Buck

Where you ever sad that you could not compute core-genomes for this collection of incomplete [MS]AGs you had lying around? Or simply to throw away a lot of data by dereplicating? I coded a thing: biorxiv.org/content/10.110… with @mehrshmali and @Stebe69 1/4

@fesshole Weaponized incompetence is a thing....

This is going to sound made up but it's not. My girlfriend told me this morning, she doesn't like the way I wash up and I'm not allowed to do it anymore. Result.

@nomad421 Brewing....

Dear #bioinformatics peeps, if you were to work in another research area (assume you could fast-track the requisite knowledge) what area would you work in, and why?

@mehrshmali @_SLU Did you also post the "one day later" pic? :P

Finally spring is here! Campus beauties at @_SLU 🤩

Apply to this long-term #postdoc position! Come and join us in our marvelous path towards understanding the role of phage co-infection in their genome evolution and become a member of the *MULTIPHAGE* #ERCStG team! Please retweet or is it reX now😉 slu.se/en/about-slu/w…

Pandoc, bless your soul.

Combinatorial sequence analysis: try the forward, then the reverse, and the rev-comp, and then repeat with R2, pick whatever matches most....

@miangoar Nah, dumber, the static layout is just based on a graph where the edges connect all genomes that have an ANI > 91.5 (slightly arbitrary, to have a reasonable picture)

@MetaMoritz Very cool viz! 😎 Which data did you use as input for Fruchterman-Reingold layout? a k-mers array for each genome? 🤔

Stuck in a species clustering rabbit hole: Pelagibacterales' ANI gap (edges ANI > 91.5%)

@K_T_Jurdzinski Well, I can't get convinced by the recombination barrier until we find a species that doesn't recombine... Anyhow, the same threshold doesn't work for some pelagi' species (they still have a barrier though) which lead me to this rabbit hole...

@K_T_Jurdzinski There is no real dimensionality reduction, it's just a standard fruchterman-reingold graph layout ^^

@naturepoker1 @biomonika In all fairness i imagine some manual copypasta scaffolding

@naturepoker1 @biomonika Then the line would be exactly twice as long though..... I would blame sequence pixies!

C'mon people. Who formats and uploads genomes to public databases like this?

@naturepoker1 @biomonika Well then i would hope the second file starts with a new entry and at least I'd see a "greater than" lost in the middle of it... Uhhh all the dirty scenarios i have in my mind now....

@m_chuvochina @AlexJProbst All smoke mirrors and more, won't doesn't matter if on 3 years postdoc you have 5 contracts, got asked to quit because the money moved faculty and told after 2.5 years grant hasn't enough money because you're not allowed to consider obligatory salary increases... #BitterPostDoc

@AlexJProbst Ah I see. To speed up the process of leaving academia 🫤 so smart…🤦‍♀️ sounds like they don’t really need scientists anymore…or maybe ‘more’ is coming?! Like creation of stable job opportunities…

@AlexJProbst @ISMEJournal @FEMSmicro I replied anyhow, but i know the story it comes out of and i think it's easy to misconstrue some aspects of this discussion without enough context.....

@AlexJProbst @ISMEJournal @FEMSmicro I do have to complain, the whole first page of questions already assumes i accept data embargoes.... Makes it difficult for me to answer without making it look like i stand behind that practice... Also, it's very black and white...

New amplicon misnomer: "Precision Metagenomics"

@BioMickWatson @Njuki_Sue @s_jaenicke All of that. But if NCBI not there the NCBI taxonomy interface of biopython can do the job, or taxize for R.

@Njuki_Sue @s_jaenicke I was super excited about the first link until I reaslised that number is NOT NCBI taxonomy ID. #harumph Thanks for this though, will explore the REST API

Bioinformatics Twitter: given a particular KEGG orthology number, I want to know: is this mostly found in Euks, mostly found in Bacteria, or found in both. How do I do this?

@BioMickWatson @npcooley The restful API is weirdly useable, between the KO entries and the ORG entries can get some info there, and can be done with `wget` and bash. But wouldn't trust it to much... It would be buased on validated stuff. I would probably run kofamscan against nr or something

@npcooley indeed it does, but then I want to take that list (sometimes 100s long) and go up to Kingdom level; and then I want to repeat for lots of KOs :)

@knstnr Fast so gut wie Tatort! ;)

Very relaxing :)