Version 0.14.0 of rapids_singlecell has been released! Super cool to see my work was able to help contribute to the addition of the wilcoxon rank-sum test to rank_genes_groups() :)
Stop spending so much time waiting for your analyses to run and start using gpu-accelerated workflows today!
hopefully PR is merged soon but it’s encouraging to see other users excited about it :)
reference != rest is still a WIP tho..
presenting a gpu-accelerated version of the wilcoxon rank-sum test for scanpy / rapids-singlecell’s rank_genes_groups tomorrow.
excited (and nervous) to show my work in a purely computational space for the first time! :)
'A functional perspective on astrocyte heterogeneity'
by Wookbong Kwon, Michael Williamson & Benjamin Deneen
www.cell.com/trends/neuro...
after the claude code integration i can say zed is the best code editor, hands down.
"Israel’s policies and actions in Gaza meet the legal definition of genocide in Article II of the United Nations Convention for the Prevention and Punishment of the Crime of Genocide (1948)".
Israel is committing genocide in Gaza, genocide scholars' association says - www.reuters.com/world/middle...
today i learned @hf.co hosts scVI models. healthy ref AstroWRLD coming soon ™️
benchmark everything :)
promote [good] research. this was a nice, quick read but has some great points that i’m going to keep in mind as i work on establishing myself in the field and to you all..
blog.slavovlab.net/2020/07/15/p...
you’re telling me I don’t have to wait 45 minutes for tl.umap() to run? …
rapids_singlecell is amazing if you’re working with large datasets
github.com/scverse/rapi...
I wrote an op-ed on the world-class STEM research ecosystem in the United States, and how this ecosystem is now under attack on multiple fronts by the current administration: newsletter.ofthebrave.org/p/im-an-awar...
Policing and punishment don't solve homelessness – they make it worse.
we need to stop anthropomorphizing llms. “model welfare” is hilarious as a concept and openai is now more focused on making chatgpt a friend rather than what it is, a tool.
the lion understands the importance of having both wet and dry lab expertise, but the lion loves to complain about wet lab
looking for collaborators in building an atlas / database of astrocyte heterogeneity in health & disease
shoutout to my hpc for enabling me to work on 1M+ cell datasets, but still, an hour to embed the umap.. time to start learning gpu integrations in the scanpy workflow
All the more power to those who don’t want to use it, but to say it doesn’t have any utility is simply untrue
A noticeable step up in code generation and organization, though. Thorough repos generated from one prompt.
Wanted to test it by having it suggest nuclei isolation strategies and it thinks I’m interested in making biological weapons or things of that nature
GPT-5 is impressive but I don’t enjoy it accusing me of violating ToS..
credentialism ruins science.
Video 2 = Drill following the path plotted by the probe.
Cranial window implantation in mice can be highly variable. This makes it less so.
Demo of a surgery robot that was adapted from www.nature.com/articles/s41...
Video 1 = Limit switch + probe to detect the surface of the skull
One must imagine Sisyphus happy (I post this after every botched astrocyte immunopanning)
theory of games and astrocyte behavior..
radar plots are nice for showing differences in astrocyte transcriptional profiles :) #rnaseq #astrocyte
