Devon Ryan

An American in Germany, an ion channel guy in bioinformatics, a proponent of scientific skepticism in an irrational world.
@[email protected]

[bioRxiv Bioinfo @biorxiv_bioinfo]

Alevin-fry-atac enables rapid and memory frugal mapping of single-cell ATAC-seq data using virtual colors for accurate genomic ... https://t.co/4cpFnTNmP3 #biorxiv_bioinfo

[Devon Ryan @dpryan79]

@BioMickWatson @nomad421 @cwvhogue @bioinformer @OmicsOmicsBlog @YouTube I don’t think they’d believe him that those are real. “Surely no one would design things this way.”

[Devon Ryan @dpryan79]

@BioMickWatson I can assure you that many of us are all too aware…

[Devon Ryan @dpryan79]

Glad they learned their lesson there. Did anyone like 0/0 for no calls?

[Stephen Turner 🦋 @stephenturner.us @strnr]

GATK 4.6.0.0 reverts back to ./. for no-calls instead of 0/0 with DP=0 https://t.co/DS12ljeWgC

[𝕐 @nomad421]

SeqKit is an absolutely amazing tool! I was writing a custom script to chunk reorder the records in a large reference file, and then chunk them into sub-files of a pre-specified size. The script wasn't huge, but was slow & error prone. Alternatively; it's just 2 seqkit commands!

[Devon Ryan @dpryan79]

@nilshomer Only a problem if you actually pay attention to your GitHub notifications…

[Heng Li @lh3lh3]

Minimap2-2.27 released with new presets for accurate Nanopore reads and Illumina Complete Long Reads (both recommended by vendors), along with a few minor features and bug fixes. Output alignment identical to v2.26 except the value at a custom tag. https://t.co/qji3vGWfoE

[Devon Ryan @dpryan79]

@yokofakun Sent via DM

[Devon Ryan @dpryan79]

I have 3 invite codes to the site where the sky is blue. Let me know if you want one.

[𝕐 @nomad421]

Very happy to see simpleaf published! @DongzeHe and I are working on some exciting improvements and expansions that I'm looking forward to sharing more about soon :).

[Dr Kareem Carr @kareem_carr]

I got featured in the Fall 2023 edition of the Harvard Public Health Magazine!

[John Marshall @jomarnz]

Do you use pysam? Currently #pysam builds wheels back to Python 3.6, which has been end-of-life for a couple of years. Please let us know via this survey which Python versions you use and would like pysam to continue supporting. https://t.co/lx70XOuVue

[Devon Ryan @dpryan79]

@nilshomer Likewise :)

[Devon Ryan @dpryan79]

Gotta say I'm feeling this at the moment. The downside to making conda packages out of anything that could prove useful is that I end up looking at the code and... yeah... well, what Nils said.

[Nils Homer @nilshomer]

Dear academics, I'm going to eventually read your code. Please feel the appropriate amount of shame from the start about your hard-coded paths, missing required files, and major assumptions about file naming. Your frenemy, industry.

[Genedata @Genedata]

#Genedata collaborates with @AstellasPharma to help confirm the #biosafety of cell therapies using #NGS-based assays within a validated environment. Read more at https://t.co/EBrqRKQmeq

#DigitalizingBiopharma #GenedataSelector #Bioinformatics #gmp #Celltherapy

[Johannes Köster (@johanneskoester.bsky.social) @johanneskoester]

@vsoch @kenbwork @PaoloDiTommaso And finally, it is great to have more than one freely available and well suitable tool to get the job done. What matters is that science is reproducible and transparent, and that these properties are strived for as many people as possible. (3/x)

[Devon Ryan @dpryan79]

@nomad421 @gtuckerkellogg @ChristosArgyrop They also have tools for type conversion, though I honestly have to say in practice this is pointing a loaded gun at your foot unless you truly have expert users.

[Devon Ryan @dpryan79]

@nomad421 @gtuckerkellogg @ChristosArgyrop For what it's worth, this is actually how workflows in Galaxy work. The individual files/groups thereof are typed and tools can be executed differently under the hood given the exact type of the inputs. They did a quite nice job with that.

[Devon Ryan @dpryan79]

I've said it before and I'll say it again, it's important that you use A workflow manager, not really which one you use. It's vital to get familiar enough with one to be productive and make reproducible and redistributable analyses. The exact choice is then totally "you do you".