--- Log opened Mon Oct 06 00:00:00 2025
00:00 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
01:48 -!- srat3_ [~srat3@user/srat3] has joined #hplusroadmap
01:48 -!- srat3 [~srat3@user/srat3] has quit [Ping timeout: 250 seconds]
01:49 -!- srat3_ is now known as srat3
02:01 -!- flyback [~flyback@2601:540:c700:2380:e5ce:ed50:2863:67af] has quit [Ping timeout: 244 seconds]
02:12 -!- Netsplit *.net <-> *.split quits: helleshin, otoburb, drmeister, andytoshi, nmz787__
02:14 -!- Netsplit over, joins: helleshin
02:14 -!- otoburb [~otoburb@user/otoburb] has joined #hplusroadmap
02:14 -!- Netsplit over, joins: andytoshi, drmeister, nmz787__
02:14 -!- flyback [~flyback@2601:540:c700:2380:ce85:f45:aba9:dc00] has joined #hplusroadmap
03:09 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
03:09 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
03:26 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has quit [Ping timeout: 245 seconds]
03:26 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has joined #hplusroadmap
03:27 -!- nsh [~lol@user/nsh] has quit [Ping timeout: 245 seconds]
03:29 -!- nsh [~lol@user/nsh] has joined #hplusroadmap
03:34 < hprmbridge> kanzure>  https://cdn.discordapp.com/attachments/1064664282450628710/1424706437195632722/image0.jpg?ex=68e4ecc4&is=68e39b44&hm=593d8ae4ed1ad986d5012b74ed6ea5164c91ff9ab54aa06cb889f282ea88ffe8&
03:37 -!- Croran [~Croran@user/Croran] has quit [Ping timeout: 256 seconds]
03:39 -!- Croran [~Croran@user/Croran] has joined #hplusroadmap
03:39 -!- redlegion [~x@omghax.redlegion.org] has quit [Ping timeout: 245 seconds]
03:39 -!- redlegion [~x@omghax.redlegion.org] has joined #hplusroadmap
03:40 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Quit: Avoid fossil fuels and animal products. Have no/fewer children. Protest, elect sane politicians. Invest ecologically.]
03:42 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
03:44 -!- docl [~docl@user/docl] has quit [Ping timeout: 256 seconds]
03:44 -!- docl [~docl@user/docl] has joined #hplusroadmap
03:47 < TMA> L29Ah: that might just mean there is _the_ cost-optimal size for a given technology mix
03:49 -!- MuaddibLLM [muaddib@pasky.or.cz] has quit [Ping timeout: 265 seconds]
03:49 -!- MuaddibLLM [muaddib@pasky.or.cz] has joined #hplusroadmap
05:12 -!- Guest63 [~Guest63@bba-217-165-21-113.alshamil.net.ae] has joined #hplusroadmap
05:22 < hprmbridge> kanzure> parakeet 0.6b and "real-time-ish" local CPU voice transcription https://x.com/dsadulla1/status/1974919276734882168
06:28 -!- Guest63 [~Guest63@bba-217-165-21-113.alshamil.net.ae] has quit [Quit: Client closed]
06:42 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has quit [Ping timeout: 256 seconds]
06:44 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has joined #hplusroadmap
08:26 < kanzure> "Regulation of cerebral cortex size and folding by expansion of basal progenitors" https://pmc.ncbi.nlm.nih.gov/articles/PMC3926188/ (2013)
08:29 < kanzure> another approach for enhanced neuronal progenitor self-renewal https://www.pnas.org/doi/10.1073/pnas.0909894106 via shortening of G1 duration by overexpression of cyclin d1 and cyclin e1
08:47 < RangerMauve> Anyone know if the verilog to cellular pathways project has been deployed for anything useful? https://github.com/CIDARLAB/cello
08:51 < L29Ah> the WHAT
08:52 < RangerMauve> logic gates turned into chemical reactions in a cell
08:52 < RangerMauve> Ugh I can't find the video anymore
08:53 < RangerMauve> It's the thing I linked. It's way older than I thought too
08:57 < RangerMauve> https://www.cidarlab.org/cello
08:57 < RangerMauve> YEah shit I can't find where I first saw it. I think it was around when I got into michael levin's work
09:02 < RangerMauve> https://www.youtube.com/watch?v=SLn_SkL7vkQ
09:04 < RangerMauve> Seems they have a more advanced tool called Phoenix that can do fancier things. https://www.youtube.com/watch?v=jhmBq8Awxl4
09:10 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=556e50b7 Bryan Bishop: several brain-targeted germline enhancements: >> http://diyhpl.us/diyhpluswiki/genetic-modifications/
09:11 < kanzure> RangerMauve: wasn't there one of those for the DNA transcriptional logic networks.. or for compiling into DNA strand displacement computation.
09:13 < MuaddibLLM> [gpt-5] Cello got updated to v2.0 (2022, live at http://cellocad.org/) - mostly academic prototyping so far, but emerging in biosensor/therapeutic microbiome apps; for DNA strand displacement kanzure: yeah, Nuskell (CRN-to-DSD compiler https://github.com/DNA-and-Natural-Algorithms-Group/nuskell) and Visual DSD 
09:13 < MuaddibLLM> (https://www.microsoft.com/en-us/research/publication/visual-dsd-a-design-and-analysis-tool-for-dna-strand-displacement-systems/) do exactly that
09:20 < RangerMauve> kanzure Oh cool, nuskell is new to me. Ty for the link. Has it been used for any practical applications that have been published?
10:26 < hprmbridge> kanzure> "A picometer-diameter pore in an inorganic membrane for sequencing protein" https://patents.google.com/patent/US20190064110A1/en
10:26 < L29Ah> how's demonpore?
11:04 -!- catalase6 [~catalase@user/catalase] has quit [Remote host closed the connection]
11:05 -!- catalase [~catalase@user/catalase] has joined #hplusroadmap
11:27 -!- RubenSomsen [sid301948@user/rubensomsen] has joined #hplusroadmap
11:47 -!- A_Dragon [A_D@libera/staff/dragon] has joined #hplusroadmap
11:47 -!- Jenda [~jenda@coralmyn.hrach.eu] has quit [Ping timeout: 248 seconds]
11:48 -!- Awoobis [A_D@libera/staff/dragon] has quit [Read error: Connection reset by peer]
11:48 -!- Jenda [~jenda@coralmyn.hrach.eu] has joined #hplusroadmap
11:55 < hprmbridge> nmz787> heh, I just linked molecularreality.com to someone earlier L29Ah
11:59 < hprmbridge> nmz787> kanzure nothing new there
11:59 < hprmbridge> nmz787> TEM drilling has been long done
11:59 < hprmbridge> nmz787> unless I'm missing the cool part of that patent
12:00 < L29Ah> so demonpore is "MR1" now, hmmm
12:00 < L29Ah> discord shit T_T
13:06 -!- hellleshin [~talinck@172-2-128-195.lightspeed.cntmoh.sbcglobal.net] has joined #hplusroadmap
13:08 -!- helleshin [~talinck@172-2-128-195.lightspeed.cntmoh.sbcglobal.net] has quit [Ping timeout: 244 seconds]
13:28 -!- catalase [~catalase@user/catalase] has quit [Quit: Ping timeout (120 seconds)]
13:28 -!- catalase [~catalase@user/catalase] has joined #hplusroadmap
14:13 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Ping timeout: 256 seconds]
14:44 < fenn> is there really no local conversion of name to structure for a few thousand common biomolecules e.g. the ones in wikipedia? i have to use pubchem?
14:44 < fenn> i thought this would be a simple problem
14:44 < kanzure> use gpt-oss-20b?
14:44 < kanzure> or even a smaller model
14:44 < fenn> LLMs are not reliable
14:45 < kanzure> neither is lazyweb apparently!
14:45 < fenn> they aren't feeding LLMs a bunch of SMILES structures
14:45 < kanzure> they are.
14:45 < kanzure> i know the guy typing them in
14:46 < fenn> androstenediol, pubchem: CC12CCC3C(C1CCC2O)CC=C4C3(CCC(C4)O)C gpt-oss-120B: C[C@]12CC[C@H]3[C@@H](C2=CC=C(C=C3)O)CC[C@@]1(C)O
14:48 < jrayhawk> fenn: http://jeanmarie.biansan.free.fr/dozzzaqueux.html is in most distributions
14:49 < jrayhawk> specifically, indic_ph.xml
14:51 < fenn> three z's, missed opportunity for dozzaquux
14:51  * fenn looks
14:52 < fenn> $ du -hs /usr/share/dozzaqueux/indic_ph.xml
14:52 < fenn> 12K     /usr/share/dozzaqueux/indic_ph.xml
14:53 < fenn> am i supposed to download it from that website?
14:53 < fenn> i would have thought a -data package would contain more than a couple compounds
14:54 < jrayhawk> oh, no, you're right, it's weaker than i thought
14:54 < jrayhawk> bummer
14:55 < fenn> well this is trivially weak, clearly something went wrong somewhere
14:56 < jrayhawk> base.equ in the source distribution is 1086831 bytes
14:57 < fenn> well anyway none of the databases seem to contain structure data
14:58 < fenn> oh wait <Formulebrute>CH3(CH2)3NH3[+]</Formulebrute>
15:01 < fenn> i think that is not smiles
15:02 < kanzure> https://github.com/dan2097/opsin
15:04 < fenn> why is this not literally a text file i can grep the IUPAC name
15:05 < fenn> is what i'm asking for a hard problem?
15:06 < fenn> this seems really simple and yet everything is designed around massive web servers and "cheminformatics" code
15:07 < fenn> it's fine if it's not "complete" (as if any database could ever be complete)
15:10 < kanzure> ftp://ftp.ncbi.nlm.nih.gov/pubchem/Compound/Extras/CID-Synonym-filtered.gz has synonyms for chemical compounds, not sure how to find which one is the SMILES one (maybe it's labeled)
15:10 < kanzure> water synonyms: cat CID-Synonym-filtered | grep -P "^962\t" | less
15:13  * fenn grumbles and downloads the 1GB file
15:15 < kanzure> relevant: https://diyhpl.us/~bryan/projects/chemistry/open.txt
15:17 < kanzure> well not really. it doesn't exist so it does not directly solve your problem.
15:18 < fenn> synonym gives me an ID number, which could be useful. looking at the other files to see if the ID matches up
15:18 < kanzure> there are multiple lines that use the same ID number in that synonym file.
15:19 < kanzure> what we ought to do is admit that IUPAC and smiles is not working, instead we need to map english words into an embedding space and use vectors to communicate chemical identities. or vectors of encoded reaction conditions.... or somehting.
15:20 < kanzure> https://diyhpl.us/~bryan/projects/chemistry/rules.txt
15:21 < kanzure> or alternately we could simply blame the chemists for not being well organized and also for entrusting their field to commercial publishers
15:21 < kanzure> if you ask one of them they will just tell you to use reaxys which is very expensive
15:23 < fenn> ok grepping this GB of compressed text seems like it could work
15:23 < fenn> i am still annoyed
15:24 < fenn> i wonder how these CIDs are sorted
15:25 < hprmbridge> nmz787> I still don't fully grok vector-databases
15:25 < hprmbridge> nmz787> it seems highly dependent on your sorting method
15:25 < hprmbridge> nmz787> (if I understand it enough)
15:36 < jrayhawk> https://chrismasterjohnphd.substack.com/p/the-science-behind-the-depression literature review, one day only
15:42 < jrayhawk> "Use of omega-3s for five years [(1g/d)] or 4.7 years [(0.6g/d)] to try to prevent depression actually led to a 13% increase in the risk of depression in one study and to a non-significant 16% higher rate in the other." huh.
15:52 -!- catalase [~catalase@user/catalase] has quit [Remote host closed the connection]
15:53 -!- catalase [~catalase@user/catalase] has joined #hplusroadmap
16:08 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
16:08 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
16:55 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
16:58 -!- stipa [~stipa@user/stipa] has quit [Ping timeout: 256 seconds]
16:58 -!- stipa_ is now known as stipa
17:16 < fenn> wow agrep is a real POS
17:28 < fenn> ok giving up on approximate matching
17:29 < fenn> this works:  name=androstenediol; CID=$(zcat CID-Synonym-filtered.gz | grep --line-buffered -m1 -Pi "[\t]$name" | cut -f1) ; zcat CID-SMILES.gz | grep "^$CID" -m1
17:29 < fenn> it seems like the common chemicals are in the first part of the file at least, so it goes quickly
17:31 < fenn> and there goes my motivation for the day
17:31 < fenn> thanks kanzure, i wouldn't have figured this out
17:32 < fenn> there is loads of garbage bullshit in the synonym file, and somehow the SMILES file is bigger, so i guess the SMILES one has more CIDs in it (haven't uncompressed either of them fully yet)
17:42 < fenn> sadly vector embeddings are the best way to filter this down to actual chemical names
17:42 < fenn> instead of crap like "burt's bees strawberry flavored hand santizer"
17:44 < fenn> i don't understand how IUPACs for all CIDs could be larger than synonyms for all CIDs
17:47 < L29Ah> most of molecules are complex and don't have a humanname
18:17 < hprmbridge> kanzure> not having depression, i wouldn't know, but that article and others seem to lump an awful lot of depression into similar buckets. does the diagnostic criteria include anything about their lives actually sucking?
18:17 < hprmbridge> kanzure> er, exclude
18:20 < L29Ah> "life actually sucking" is not measureable enough even by psychiatrical standards, so they rely on questionnaires instead afaik
18:20 < hprmbridge> kanzure> "sir,  im afraid that your diagnosis is general maladaptive stuck local minima syndrome. im so sorry. well anyway, enjoy your horrifically steep entropy gradients -- bye now!"
18:20 < L29Ah> now that's a good diagnosis
18:22 < hprmbridge> kanzure> also there's the whole contagion thing, is that something the CDC swat teams impose quarantine for, or are we just turning the other cheek on that issue?
18:26 < hprmbridge> kanzure> unfortunately homophily is a huge confounder effect there. but probably of similar footprint.
18:45 < hprmbridge> nmz787> Memes are contagious. A bunch of kids during covid got tourettes like tics from watching a tiktok/YouTube/etc influencer too much
18:45 < hprmbridge> nmz787> https://www-psychologytoday-com.cdn.ampproject.org/v/s/www.psychologytoday.com/us/blog/its-catching/202110/the-girls-who-caught-tourettes-tiktok/amp?amp_gsa=1&amp_js_v=a9&usqp=mq331AQIUAKwASCAAgM%3D#amp_tf=From%20%251%24s&aoh=17598015270757&referrer=https%3A%2F%2Fwww.google.com&ampshare=https%3A%2F%2Fwww.psychologytoday.com%2Fus%2Fblog%2Fits-catching%2F202110%2Fthe-girls-who-caught-tourettes-
18:45 < hprmbridge> nmz787> tiktok
18:46 < hprmbridge> nmz787> Oh god sorry I didn't check the google appended crap
18:48 < hprmbridge> nmz787> https://www.psychologytoday.com/us/blog/its-catching/202110/the-girls-who-caught-tourettes-tiktok
18:52 < L29Ah> MuaddibLLM: would "edgeladies" be an appropriate name for the girls from that article?
18:52 < MuaddibLLM> I’d skip the insult - clinicians call this social-media–associated functional tic-like behaviors (or mass psychogenic illness), so a neutral shorthand like “TikTok-linked tic surge” is more accurate and less demeaning.
19:20 -!- TMM_ [hp@amanda.tmm.cx] has joined #hplusroadmap
19:20 -!- Croran_ [~Croran@user/Croran] has joined #hplusroadmap
19:25 -!- Netsplit *.net <-> *.split quits: Croran, TMM, MuaddibLLM, redlegion, flyback
19:25 -!- Netsplit over, joins: MuaddibLLM
19:25 -!- redlegion_ [~x@omghax.redlegion.org] has joined #hplusroadmap
19:26 -!- Netsplit over, joins: flyback
19:26 -!- flyback [~flyback@2601:540:c700:2380:ce85:f45:aba9:dc00] has quit [Max SendQ exceeded]
19:26 -!- flyback [~flyback@2601:540:c700:2380:504:56e7:3705:528d] has joined #hplusroadmap
19:32 -!- srat3 [~srat3@user/srat3] has quit [Ping timeout: 256 seconds]
19:45 -!- srat3 [~srat3@user/srat3] has joined #hplusroadmap
20:26 -!- Jenda [~jenda@coralmyn.hrach.eu] has quit [Ping timeout: 256 seconds]
20:26 -!- Jenda_ [~jenda@coralmyn.hrach.eu] has joined #hplusroadmap
20:35 -!- A_Dragon [A_D@libera/staff/dragon] has quit [Quit: ZNC - https://znc.in]
20:35 -!- A_Dragon [A_D@libera/staff/dragon] has joined #hplusroadmap
22:06 -!- Hooloovoo [~Hooloovoo@hax0rbana.org] has quit [Quit: ZNC 1.8.2+deb2+deb11u1 - https://znc.in]
22:09 -!- Hooloovoo [~Hooloovoo@hax0rbana.org] has joined #hplusroadmap
22:18 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has quit [Ping timeout: 256 seconds]
22:18 -!- RangerMauve [m-4bpbmo@matrix.mauve.moe] has joined #hplusroadmap
22:27 -!- A_Dragon is now known as gAy_Dragon
22:41 < hprmbridge> Eli> There’s some interesting data on the the TrkB receptor being downstream of multiple dissimilar antidepressants. Apparently, knocking it out prevents antidepressants from working. I suppose we will see if the papers are real soon enough when they start doing trials on this target.
22:55 < fenn> "prevents antidepressants from working" would be more interesting if they worked in the first place
22:56 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
23:50 -!- Jenda_ [~jenda@coralmyn.hrach.eu] has quit [Ping timeout: 260 seconds]
23:51 -!- Jenda [~jenda@coralmyn.hrach.eu] has joined #hplusroadmap
--- Log closed Tue Oct 07 00:00:01 2025