Daniel Cook ∛

DeepSomatic (github.com/google/deepsom…) preprint is out showing improvements in somatic variant calling in various platforms. Work lead by @jiimiinpaark and @daniel_e_cook. Tumor-only lead by @pichuan. In collaboration with @acarroll_ATG, @MishaKolmogorov and @BenedictPaten. 🧵1/7

@T1MD1 @fulcrumgenomics Congrats Tim - sounds like a great new role!

I'm thrilled to announce that I successfully defended my PhD thesis today, "Improving select applications of long-read DNA sequencing"! After graduating, I'll be joining @fulcrumgenomics as a Senior Bioinformatics Scientist. I can't wait to see what this next chapter brings!

@d0m96 Alfredapp.com works well. Supports more complex operations (log, exp, etc) if you prefix your calculation with ‘=‘. Has a history and clipboard support.

How do you do simple calculations on your computer? Is there anything faster than pasting them into Google?

@XLR For me, it’s 0.38 or nothing! jetpens.com/Uni-ball-Signo…

My old 0.38 felt a bit too sharp for note taking. Got some new MUJI pens, couldn't be happier

@daniel_c0deb0t @AnthropicAI Congrats and best of luck in your new role! I look forward to seeing what you will do next!!

Bit of a plot twist, but I'll be joining @AnthropicAI after I graduate undergrad at UCLA to work on LLM research tools

@daniel_c0deb0t latex? I had this problem recently. It took me at least 4 hours to figure out.

@aarondfrancis There are two sets of twins in my family. One identical and one fraternal. It’s fun because I can say I have 3 twin brothers. Good luck!

🥳🥳🥳 I'm incredibly excited to share that my wife and I are expecting twins in the fall! You may be thinking... but Aaron, you already have twins. Yes. Yes, we do. We're going to have another set of twins, somehow. Four kids under three years old. Send help!

@zamanian_ Congrats!!!

@shaneenz28 Congrats!!!

I will forever be the girl who gets too into the theme. But I married someone who lets me do makeup on him so I’m fine with it

@karpathy I prefer: `import IPython; IPython.embed(user_ns=dict(globals(), **locals()))` Which gives an ipython interpreter.

debugging in Python: - `print()`s alone: too simple - `import pdb; pdb.set_trace()`: too complex - `import code; code.interact(local=locals())`: just right simply drops you into interpreter, perfect for 95% of debugging

@aaronquinlan what the golf? is fun!

Is there a single good game available on Apple Arcade?

This @rustlang tool is part of my work @GoogleHealth’s genomics team (hosts: @daniel_e_cook and @belyaeva_ana). We use it for evaluating DeepConsensus, though it should be generally useful for evaluating basecalling or other sequencing methods Github: github.com/google/best

@ChiBuildings Is it a wood fired pizza place?

@vsbuffalo @BioMickWatson @pmelsted @GuillaumOleSan Laughed so hard at this @vsbuffalo

@BioMickWatson @pmelsted @GuillaumOleSan Yes, Jesus is 1-based too since there is no year zero between 1 BC and 1 AD en.wikipedia.org/wiki/Year_zero.

@ChiBuildings “A more welcoming atmosphere.” Not sure the layout was the issue 😆

@brent_p It can be helpful to record software versions to a file when using many conda environments for data provenance purposes.

@brent_p @brent_p This is the approach I take generally. I tried to create a general purpose environment but too many dependencies clashed. This works well in the context of Nextflow, and I also use conda within Docker for convenience.

in summary. the resolver can't handle many dependencies (or many are just broken, i don't care which), so you must use many envs to separate dependencies (semi) manually. and that's where we're at. do i get that right?

@brent_p @brent_p Looks great! I like to allow users to set the output dir. e.g. publishDir "${params.output_dir}/gvcfs"

first release of new rare-disease workflow project. it's early days, but if you are familiar with nextflow and have rare-disease cohorts, you can get high-quality SNP/indel candidates from *.cram/bam with a single command. github.com/brentp/rare-di… feedback welcome.