Erik Garrison

Erik Garrison

@thinks.lol

computational biologist: genomes, evolution, DNA

2.4K followers73 following30 posts

Erik Garrison reposted

Alex Crits-Christoph
@acritschristoph.bsky.social

Replying to @carlzimmer.com

see also this work, which helps provide some helpful framing of how novel these phages are: www.biorxiv.org/content/10.6... tl;dr: I believe this work generates only marginally novel phages, which have only a moderate likelihood of working, that decreases the more novel they get.

biorxiv.orgQuantifying evolutionary novelty and design efficiency in generative genome designGenerative genome design models can now produce previously unobserved genome-length sequences, but assessing their capabilities is complicated by limitations in functional prediction. The ability to e...

Erik Garrison reposted

Rob Edwards
@linsalrob.bsky.social

Replying to @acritschristoph.bsky.social

This is a much more impressive reimagination of phiX174, which, alas, didn't get an NYTimes article about it because of, well, the other used AI. www.biorxiv.org/content/10.6...

biorxiv.orgComplete Mutagenesis of the Genome and Proteome of ΦX174The bacteriophage ΦX174 was the first genome to be sequenced and the first to be chemically synthesised. Here we present a complete map of the consequences of changing every nucleotide in the ΦX174 ge...
Erik Garrison
@thinks.lol

Postdoc position opening in my group! Research projects: pangenomes for diverse organisms, genome evolution, biocomputing, language models. Please reach out if interested!

Erik Garrison
@thinks.lol

We've had a wonderful series of pangenome meetings in Memphis. Join us for the next edition in May!

Andrea Guarracino@andreaguarracino.bsky.social

Want to level up in #pangenomics? Join our workshop, conference & biohackathon in #Memphis, May 11-15, 2025. Connect with leading scientists, embrace genomic diversity, and contribute to cutting-edge software. Register now! pangenome.github.io/MemPanG25/ #Bioinformatics #MemPanG25

Erik Garrison
@thinks.lol

Speech to text is the most efficient way to communicate, but you have to learn how to do it, just like you learn how to touch type or tap on your phone. And it's not just writing, but reading that we have to learn how to do, because we're not used to reading speech-to-text transcriptions.

View full thread
Erik Garrison
@thinks.lol

I'm willing to bet that that kind of environment doesn't look a whole lot like Blue Sky or Twitter, where you have only 300 characters to say what you mean to say. What you realize when you start speaking to type is that that can be filled in seconds.

Erik Garrison
@thinks.lol

Speech to text is the most efficient way to communicate, but you have to learn how to do it, just like you learn how to touch type or tap on your phone. And it's not just writing, but reading that we have to learn how to do, because we're not used to reading speech-to-text transcriptions.

View full thread
Erik Garrison
@thinks.lol

We can put our thoughts into words in a dynamic flow, and it's also something you can learn to do better. It's not something we're practicing as much now because we sub-vocalize ideas into text using our fingers, and then we go back and edit them.

Erik Garrison
@thinks.lol

But where are people practicing together, not just alone talking with machines or using this as a faster way to type, but really in a group setting where people are encouraged to speak their mind without adulteration and to directly flow their thoughts into a collective consciousness.

Erik Garrison
@thinks.lol

Speech to text is the most efficient way to communicate, but you have to learn how to do it, just like you learn how to touch type or tap on your phone. And it's not just writing, but reading that we have to learn how to do, because we're not used to reading speech-to-text transcriptions.

View full thread
Erik Garrison
@thinks.lol

The biggest problem in my experience is that actually the recipients have trouble parsing what's been said because so much can be said so quickly like dramatically faster than you can write

Erik Garrison
@thinks.lol

It's not just the raw words per minute of the typing versus speaking that matters. It's that we're dramatically better at speaking. It is what we've evolved to do for millions of years.

Erik Garrison
@thinks.lol

Speech to text is the most efficient way to communicate, but you have to learn how to do it, just like you learn how to touch type or tap on your phone. And it's not just writing, but reading that we have to learn how to do, because we're not used to reading speech-to-text transcriptions.

Erik Garrison
@thinks.lol

I'm really fascinated by the dynamics that might arise in a group chat or asynchronous messaging system where everyone is encouraged to use speech-to-text and to do minimal correction of what they write.

Erik Garrison
@thinks.lol

You could set this up a few ways. You could have a group chat where, by convention, people are behaving this way. Or you could have a kind of application that makes it seamless for the user to do this.

Erik Garrison reposted

Adam Phillippy
@aphillippy.bsky.social

Kateryna Makova and I were excited to be guests on Science Friday to discuss the Y chromosome. Congrats to the amazing @arhie.bsky.social, Sergey Nurk, and others on completing this last unfinished chromosome! Tune in here if you’re curious about ChrY: www.sciencefriday.com/segments/y-c...

sciencefriday.comUnraveling the Mysteries Of The Y ChromosomeAssembling the complete sequence of the Y chromosome is an important step toward understanding the human genome.
Erik Garrison
@thinks.lol

We are already using this at scale to build an all-to-all homology map for the vegetate genomes project 290 genome set! It let us find similarities at down to 70% ANI in "only" ~200k CPU hours. Was totally impossible with minimizer mashmap

Adam Phillippy@aphillippy.bsky.social

"Minmers are a generalization of minimizers that enable unbiased local jaccard estimation" aka MashMap3 with Bryce Kille, @erikg.bsky.social, and @treangen.bsky.social is now accepted in Bioinformatics! Many thanks to the reviewers for their very careful reviews doi.org/10.1093/bioi...

You are up to date.