@merm_bot You can't really use those for complex pipelines, and the issue is that most biological software is abandonware or is fussy about dependencies, conda makes it easier.
Alongside Wisteria, I'm also excited to introduce another accelerated bioinformatics tool, Sedum, which you can use to estimate sequencing coverage and depth up to two times faster than mosdepth, while using 90% less memory per run.
You can try it here: https://t.co/9RBKlRk33z
I am happy to introduce Wisteria, an alternative over 12x faster than FastQC, developed especially for long reads with native BAM support and chart generation.
Have a look at it here: https://t.co/n9sCDbUL0T
@DVirilis It's becoming clear now that they are starting to push researchers out of their platforms to start running the pipelines internally. After all if they make the discoveries they can capture far more value than selling tokens to researchers.
@DVirilis I agree. And I think that if an institutional bad actor would develop such a pathogen they would have compute and resources, and be Independent of Anthropic's models.
@TensorTwerker It's clear that they want to gatekeep researchers from using their models since a researcher can create higher value than the one they can capture by selling tokens. It is also quite sad that to do this they claim that everyone can make a bioweapon in an afternoon.
@BapnaArihant Congrats on the round! Also happy to see startups with such polished branding and UX from day one. I suggest you look into biology as well with this, sequencing workflows especially generate large amounts of data and this would help a lot with that.
@brad_woolf@nanopore Interestingly, Chinese company BGI has mined deep seam metagenomes for novel nanopores, landing a CsgG variant with little similarity to existing sequences while maintaining the precise scaffold. They still had to engineer it to tune pore current and translocation speed.
@brad_woolf@nanopore You can either start from a wild-type scaffold and engineer or directly find candidates that match in nature, although almost every candidate will require some level of engineering to work properly. They seem to have filed for a patent with this exact nanopore.
@brad_woolf@nanopore I presume they engineered it to have a narrower constriction at both construction sites (or mined for one). Together with CsgG secretins are structurally conserved and are incredibly stable once inserted in a membrane.
@brad_woolf@nanopore Actually, they seem to use something closer to the Vibrio cholerae Type II Secretion system protein (RCSB: 9K8V), you can see it in the picture below, compared to CsgG.
@PetrovichBilly Yeah it can still happen unfortunately, the plant was just in stasys and now realised it's not really it's thing to live anymore, happened to me too
@PetrovichBilly Yeah because selections are a way to market plans for higher prices as usually the whole point is that it's a special version of the regular species. However, taxonomically it simply does not exist. You can keep it on the tag though! And should actually