IRC channel logs

2022-12-20.log

back to list of logs

<zimoun>hi!
<civodul>hey, zimoun!
<rekado>need to share a good joke
<rekado> https://github.com/plotly/Kaleido
<rekado>“Fast static image export for web-based visualization libraries with zero dependencies”
<rekado>“To accomplish this goal, Kaleido introduces a new approach. The core of Kaleido is a standalone C++ application that embeds the open-source Chromium browser as a library.”
<rekado>¯\_(ツ)_/¯
<civodul>ah ah, fun :-)
<civodul>it should include the open-source Emscripten as well, so it can compile open things to wasm to run in Chromium
<rekado>great idea! Pull request?
<efraim>I tried to build kaleido months ago, I ended up introducing an inferior into a channel to try to pull out a specific chromium version. it was a huge mess
<civodul>seriously? that's surprising because, hey, "zero dependencies"!
<rekado>I wanted to build https://github.com/aertslab/scenicplus/blob/main/requirements.txt
<rekado>it requires kaleido and ray
<rekado>both of these would probably take me a month or more to package
<civodul>a use case for "guix shell -CF"? :-)
<zimoun>Last Friday, I attended as a spectator to Master bioinfo projects. The idea is to give to students a paper at the beginning of the semester and ask them to reproduce some parts and share their findings.
<zimoun>This year, the paper was https://doi.org/10.1158/2159-8290.CD-13-0330 from 2013.
<zimoun>
<zimoun>Very interesting. 11 groups of 3 or 4 students, none finds the same conclusion.
<zimoun>The paper concludes about a list of genes for Cancer.
<zimoun>Using the exact same data, the students are not able to find the same list; or at least some genes from this list.
<zimoun>The tools are different; for instance genomic aligner, R package for computing gene experession, etc.
<zimoun>Well, I have not yet confirmed by my hypothesis about this difference. I guess it comes from the genomic reference (hg19 or earlier vs hg38).
<zimoun>rekado: is pigx able to run RNA-seq differential analysis?
<zimoun>it does. :-)
<rekado>zimoun: yes, you have to add the DE analyses configuration to the settings file to enable it.
<rekado>efraim: do you have a WIP package definition for kaleido?
<efraim>rekado: yeah, let me try to find the link on gitlab
<efraim> https://gitlab.com/genenetwork/guix-bioinformatics/-/commit/a57961ff6192bf33f2c9c1fae995f3b2ef620c87
<efraim>that was the last time I touched it
<efraim>I had to move it to its own branch after that and never looked at it again
<civodul>zimoun: interesting exercise, and concerning too!
<zimoun>civodul, yeah. I was overthere because I would to redo using Guix (instead of Docker, blabla)
<civodul>of course :-)
<civodul>thing is, it's concerning if people aren't able to get the same results on a cancer study, using the same input data
<civodul>not that i'm really surprised, but it tells a lot
<civodul>when people "out there" will realize the mess scientists are in, we'll be in trouble