--- Log opened Sat Apr 11 00:00:49 2020 00:01 -!- filipepe [uid362247@gateway/web/irccloud.com/x-urghsgdvcefcetof] has quit [Quit: Connection closed for inactivity] 00:39 -!- darsie [~kvirc@84-114-73-160.cable.dynamic.surfer.at] has joined ##hplusroadmap 01:40 -!- Human_G33k [~HumanG33k@62.147.242.8] has joined ##hplusroadmap 01:43 -!- HumanGeek [~HumanG33k@62.147.242.8] has quit [Ping timeout: 250 seconds] 02:45 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:712e:945c:72a9:6455] has joined ##hplusroadmap 02:46 < nmz787> fenn: https://visormaskshield.com/ 02:48 < nmz787> made by the company I work part-time for 02:51 < fenn> cool, but not a substitute for a respirator 02:51 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:712e:945c:72a9:6455] has quit [] 02:51 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:712e:945c:72a9:6455] has joined ##hplusroadmap 02:53 < fenn> i'd like to see more DIY and startup PAPRs 02:53 < fenn> it'd be neat if you could run it on USB power 02:54 < fenn> by "more" i mean i haven't seen any at all 02:54 < fenn> more than zero 02:57 < srk> this one is close I think https://github.com/jcl5m1/ventilator 03:01 < fenn> huh? he seems confused what PAPR means 03:01 < fenn> i don't understand why it's even in this project in the first place 03:02 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:712e:945c:72a9:6455] has quit [] 03:03 < srk> well at first he wanted low-cost oss ventilator but realized it's not that easy 03:03 < fenn> so make a new repository 03:03 < srk> so the project changed a bit 03:05 < srk> he should have created an org instead and try to organize people 03:05 < fenn> "These adapter plates can be attached with liberal amounts of hot glue" 03:05 < fenn> umm, no 03:06 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:712e:945c:72a9:6455] has joined ##hplusroadmap 03:14 < fenn> i would feel bad for wasting someone's time if i recommended this project 03:14 < srk> well yeah, most of these projects are like that sadly 03:15 < fenn> where are all the recently idle mechanical engineers 03:16 < srk> so what's acutally needed for good PAPR? 03:17 < fenn> not a whole lot. a blower that has the right parameters, a filter, airtight seals between all the parts (sadly lacking in the previous project), a flexible hose, a hood with a face shield 03:18 < fenn> oh and a battery and some way to clip the whole assemblage to your belt 03:19 < fenn> it should work with 3M P100 filter cartridges and USB batteries for maximum interoperability 03:19 < fenn> ideally all the parts could be 3d printed and/or injection molded 03:19 < fenn> or at least resin cast 03:20 < srk> thanks! 03:20 < fenn> i was going to design mine to use a dewalt battery pack, when i needed something like this for woodworking 03:20 < srk> was about to say that USB batteries sound weird 03:20 < fenn> the dewalt batteries are a lot more expensive than a scrapped laptop battery pack 03:21 < fenn> they make these 18650 "power bank" things that can be ok 03:21 < srk> yeah,RC LiPos are cheap but dangerous a bit 03:21 < fenn> no don't use a LiPo 03:21 < srk> that's just 18650 with chargin circuitry and step-up converter 03:21 < fenn> yeah 03:22 < fenn> anyway i'm sure people have a lot of stuff that could work 03:22 < srk> could work if you pick one with like 2A of current 03:22 < fenn> it's not clear to me what the actual power requirements are 03:22 < srk> (LiFePo4 would be better) 03:22 < srk> I think on that page he says how much that blower needs 03:22 < srk> I think it's somethin like 10-20W 03:23 < fenn> the random blower he found behind his toilet 03:23 < srk> no that's from CPAP machine :) 03:23 < fenn> right, like i said 03:24 < srk> related https://drmrehorst.blogspot.com/2018/04/the-mother-of-all-print-cooling-fans.html 03:24 < srk> quote from there "Yup, about 1A which works out to about 24W. The CPAP blower uses about 12W and it moves about the same amount of air." 03:25 < fenn> can a hard disk drive run at 24W continuous? 03:25 < fenn> i've seen a few printed blowers based on CPU case fans 03:26 < fenn> that blower design looks wrong 03:27 < fenn> there are huge gaps above and below it 03:27 < srk> he discusses that in the video IIRC (continuosly running at 24W), they are not rated and he didn't test that for much long 03:28 < srk> but I like the idea of using HDD motor 03:28 < fenn> and the fan outlet is on the wrong Z level? 03:28 < fenn> i mean on the yellow version 03:28 < fenn> anyway i'm done critiquing designs 03:29 < fenn> have some more complaining, courtesy of Dr. Zayner: http://theantisense.com/2020/04/01/all-my-friends-are-dying/ 03:29 < srk> indeed 03:30 < srk> such blower design is interesting as it looks like countering common sense fin-orientation wise 03:33 < srk> ( https://youtu.be/IiE8skW8btE?t=301 ) 03:33 < fenn> how come heroin junkies aren't all dying of toxic shock/pyrogens 03:48 < fltrz> so the heme paper I mentioned looks like total bullshit on closer inspection: proudly report a *humongous* binding energy, I believe they thought positive and higher binding energy meant higher binding affinity *facepalm* 03:49 < fltrz> not even sure how their software allows itself to be configured to do this 03:50 < fltrz> "Discovery Studio" 03:55 < fenn> "In the German language, there is no word for “engineering.” The public says words such as “genetically manipulated organisms,” “this is genetically manipulated” or “genetically changed or altered.” The word “manipulation” has an especially strong meaning, as it implies foul play in German. For example, elections can be manipulated." 04:02 < fenn> more complaining http://theantisense.com/2018/11/12/the-people-who-fall-through-the-cracks/ 04:02 < fenn> how is it there isn't a single country in the world that can help these people 04:27 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap 04:54 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:712e:945c:72a9:6455] has quit [Remote host closed the connection] 05:01 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:712e:945c:72a9:6455] has joined ##hplusroadmap 05:45 < fltrz> erm Ingenieur(in) is german for engineer... 05:47 < fltrz> "engineering" is Ingenieurwesen 05:48 < fltrz> "Engineered" => entwickelt 06:25 < fltrz> when RNA is transated into protein, say it has already chain the first N out M amino acids, can the partial protein that is still being described already start folding? are there short sequences in the primary structure that temporarily bind to spots on the ribosome, until a certain condition is met (perhaps tug strengths rise with length of the loop? perhaps a matching later subsequence binds to the previously bound subsequence?) i.e. can we learn 06:25 < fltrz> more about folding as translation happens?) 06:26 -!- filipepe [uid362247@gateway/web/irccloud.com/x-ffifwdoipcesblef] has joined ##hplusroadmap 06:27 < fltrz> for a short sequence the space of possible folds is much smaller, so if it gets in the correct folding structure as it is being translated, then it implies folding search space remains small as amino acids are added incrementally 06:30 < fltrz> instead of an exponential search space it would mean a roughly linear one 06:31 < TMA> good heuristic to try out 06:35 < fltrz> TMA, so how would one go about training a ML algorithm? I think I would need a database of proteins PDB files? then I need to know which end of the protein was translated first, then essentialy consider protein structures at step N as being the PDB with all amino acids after N removed. then train an algorithm on folding the next X amino acids? 06:37 < fltrz> I'm not too familiar with details, so I don't know which end would be the start. are the amino acids translated from RNA in the same order the DNA was transcribed to RNA? or is it in reverse order? 06:38 < TMA> as a first aproximation -- I think you shall allow to unfold a folded part up to a certain lenght 06:39 < fltrz> TMA so thats effectively the undecided length of amino acids connecting the correctly folded partial protein with the ribosome still translating RNA? 06:40 < TMA> fltrz: sorry, I know probably less than you. it just sounded like something that could work and that should at least be tested 06:40 < fltrz> not sure you know less than me though ;) 06:41 * TMA just recognizes some of the keywords 06:41 < fltrz> it just always seemed odd that physical time should correspond with "textbook time" i.e. step 1) transcribe 2) translate 3) fold 06:41 < fltrz> but I don't see why folding of the partial protein couldn't be happening during translation 06:42 < TMA> from what I have seen the ribosome is a blob with a hole, that excretes the amino chain 06:42 < fltrz> yep, and the excretion is typically drawn like a random amino acid chain... but I question the veracity of this educational drawing 06:43 < TMA> as soon as the end of that chain is far enough from the hole as not to be blocked by the hole physics happen so that the chain begins to fold IMHO 06:43 < fltrz> thats exactly what I suspect too 06:43 < TMA> because there is no mechanism for it to stop happening -- there is 06:43 < TMA> "environment 06:44 < TMA> " all around 06:44 < TMA> [" is near return and I am clumsy, sorry] 06:45 < TMA> I don't even know, how the current folders operate. 06:46 < fltrz> if it is true it should mean the following, if we train a fixed algorithm, but train 2 different instances: one on sequences and protein structures in the forward way, and the other in the backward way; then if this shorter incremental folding search space is biologically exploited, then one algorithm trained in the correct sense should significantly outperform the other; while if it is not biologically exploited, both should be as poor as each 06:46 < fltrz> other at the task 06:47 < fltrz> I believe this should even show up as a significant difference in prediction error rate if the trained folder is very suboptimal 06:47 < fltrz> so it sounds rather falsifiable, hence interesting to try 06:59 < fltrz> so the entropic force on a chain in a thermal bath increases with length, so if it folds correctly at short length then, the length between partial protein and ribosome is relatively short and the tug force is low, while if it fails to fold for a longer length the entropic thermal tug forces become larger so a misfold on the protein get folded loose 06:59 < fltrz> interesting 07:01 < fltrz> so essentially model the ribosome as a point mass with the same weight as the ribosome, and simulate how the protein folds while lengthening the chain of amino acids incrementally 07:01 < fltrz> in a solvent with a thermostat 07:03 < fltrz> this ignores any temporary "handles" on the ribosome surface 07:04 -!- CryptoDavid [uid14990@gateway/web/irccloud.com/x-qkjmdpkxkfmikcgk] has joined ##hplusroadmap 07:11 < TMA> I think that the electrical field of the ribosome needs to be modelled as well 07:16 < fltrz> yes, I think it would be modeled by quantum chemistry modelers 07:19 < kanzure> fltrz: yes, proteins can partially fold while the ribosome is still doing it sowrk 07:19 < fltrz> I wonder if there are hydrophobic pockets on the ribosome surface 07:19 < kanzure> *its work 07:20 < fltrz> kanzure, sounds super interesting, does it have a name or where can I learn more? 07:23 < fltrz> kanzure, if you don't have time now that's ok, but I would really love to find out more about this :) 07:28 < kanzure> i think it's called cotranslation 07:29 < kanzure> .title https://www.biorxiv.org/content/10.1101/722611v2.full 07:29 < saxo> Visualising nascent chain dynamics at the ribosome exit tunnel by cryo-electron microscopy | bioRxiv 07:30 < kanzure> fenn: jrayhawk: thoughts/comments? https://github.com/kanzure/diyhpluswiki/pull/92/files 07:33 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=efb34ae8 Michael Folkson: Add link to Simplicity mailing list >> http://diyhpl.us/diyhpluswiki/transcripts/blockstream-webinars/2020-04-08-adam-back-simplicity/ 07:33 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=e485f85d Michael Folkson: Merge pull request #91 from kanzure/add-mailing-list >> http://diyhpl.us/diyhpluswiki/ 07:33 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=afa98a6a Bryan Bishop: fix typo >> http://diyhpl.us/diyhpluswiki/transcripts/blockstream-webinars/2020-04-08-adam-back-simplicity/ 07:37 < fltrz> kanzure, thank you very much 07:55 < fltrz> hm whats a "linker" ? "Initial analyses of these RNCs revealed the transition of FLN5 to the folded state occurs when it is bound to the ribosome by a linker of approximately 45 amino acids" 07:58 < fltrz> so its clear from this paper that one would have to model this exit tunnel of the ribosome and how the "nascent chain" interacts with it 08:10 -!- yashgaroth [~ffffffff@172.58.19.12] has joined ##hplusroadmap 08:24 < fltrz> huh, https://www.pnas.org/content/117/7/3528 08:25 < fltrz> so even though the translated chain is identical, one case can fold worse depending on what synonym was in the RNA transcript? 08:25 * fltrz 's brain melts 08:35 < fltrz> "In recent years, however, it has become clear that synonymous substitutions can significantly alter protein function in vivo through a wide variety of mechanisms that can change protein level (3–5), translational accuracy (6, 7), secretion efficiency (8, 9), the final folded structure (1, 10–12), and posttranslational modifications (13)." 08:35 -!- filipepe [uid362247@gateway/web/irccloud.com/x-ffifwdoipcesblef] has quit [Quit: Connection closed for inactivity] 08:37 < fenn> kanzure: what about non-bitcoin transcripts 08:37 < fltrz> how can 2 identical amino acid chains have different excretion efficiencies or post-translational modifications if they derive from different RNA's? 08:38 < fltrz> unless the ribosome does something to the "same" amino acids referred to by "synonymous" codons 08:39 < fltrz> if this is real, then it's not very surprising that folding is so hard if the "synonymous" codons are in fact not synonymous and code for modified amino acids 08:40 < fenn> flrtz i havent read your whole wall of text yet, but "can a partially transcribed protein fold?" the answer is yes, and often it folds incorrectly so you need "chaperone" proteins to stop that from happening 08:41 < fltrz> fenn, yes thank you, kanzure already helped out with a reference so now I have some keywords thanks to that reference. so the chaperons already help folding while the protein is only starting to get born :) 08:42 < fltrz> the ribosome and chaperons play this complicated collaborative role in folding during translation 08:43 < fltrz> I am totally ready to accept that, but now I'm stumped reading about how "synonymous" codons have non-synonymous differences, but how can that be unless they aren't truly synonymous? 08:43 < fenn> i think the chaperones mostly just prevent folding 08:43 < fltrz> fenn, oh, ok, this all new for me :) 08:44 < fltrz> how does a chaperon know where to bind on the amino sequence? or natural selection places "chaperone binding sites" in the protein sequence? 08:45 < fenn> it doesn't bind the new protein, it just makes a narrow tunnel through which it is extruded 08:46 < fltrz> fenn, the chaperones make a narrow tunnel? 08:46 < fenn> that's what it looks like. i'm sort of making this up 08:47 < fltrz> this is a different external tunnel, or it is the same tunnel as the ribosome tunnel? 08:48 < fenn> the ribosome exit tunnel is very short 08:48 < fenn> it probably exists for the same reason 08:49 < fenn> imagine the chaperones are just an extension of that tunnel 08:49 < fltrz> fenn, aha 08:49 < fltrz> one could almost say those chaperones are an unrecognized subunit of the ribosome then? 08:50 < fenn> one could almost say a lot of things 08:50 < fltrz> :) 08:50 < fltrz> do you have any clue how the synonymous codons can produce those weird effects after translation to the "same" amino acid sequence? 08:52 < fltrz> is it in fact the same amino acid sequence, or is that a mass propagated error in text books? I was always told the synonymous codons produce the same amino acid 08:53 < fenn> what weird effects 08:54 < kanzure> fenn: i don't know what to do about non-bitcoin transcripts. maybe move them too? i don't know. 08:54 < fenn> kanzure i'd suggest moving only the bitcoin transcripts. this means you'll need a lot of individual redirect rules 08:54 < fenn> its like 50% bitcoin transcripts 08:55 < fltrz> fenn, the weird effect: https://www.pnas.org/content/117/7/3528 ""In recent years, however, it has become clear that synonymous substitutions can significantly alter protein function in vivo through a wide variety of mechanisms that can change protein level (3–5), translational accuracy (6, 7), secretion efficiency (8, 9), the final folded structure (1, 10–12), and posttranslational modifications (13)." 08:56 < fenn> for protein level, accuracy, and efficiency, that's because there are different levels of each codon tRNA in a given cell 08:56 < fltrz> while the textbook interpretation is that since synonymous codons map to identical amino acids, they will be identical proteins with identical properties 08:56 < fenn> if you're limited by a particular tRNA then the protein will have to wait until it finds one, so it will be slower, or it may just pick the wrong tRNA while it's waiting 08:58 < fenn> i guess you could explain posttranslational modifications the same way, where if it's halfway done and dangling around still stuck to the ribosome, it has exposed bits that normally aren't exposed, and those bits can be modified 08:59 < fenn> er, some of those "it"s referred to a ribosome and some to the protein 08:59 < fltrz> fenn, I agree the only differences must occur during translation... 09:00 < fenn> there could theoretically be modified amino acids that have their own codons 09:00 < fenn> somebody would have noticed by now 09:00 < fltrz> but I was wondering perhaps the ribosome modifies the amino acid subtly depending on "synonym" variant 09:00 < fltrz> isomer or is that impossible? 09:01 < fenn> there are steric amino acids, sure 09:01 < fenn> it's not impossible, just not observed 09:01 < fltrz> hmmm 09:02 < fltrz> perhaps this paper and others like it are starting to observe it? 09:03 < fenn> sorry i'm not going to read the paper, it's well past time for bed 09:03 < fltrz> yeah, your comments were helpful, thanks! 09:05 < fenn> cotranslation is just multiple ribosomes on the same mRNA, like beads on a string 09:11 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:712e:945c:72a9:6455] has quit [] 09:14 < andytoshi> .title https://www.tillett.info/2020/04/05/a-solution-to-covid-19/ 09:14 < saxo> A (possible) solution to COVID-19 – Daniel Tillett 09:15 < andytoshi> tl;dr find some mutant sars-cov-2 strain which is not very deadly, deliberately spread it around, the fda can't stop this 09:21 < yashgaroth> fltrz the role of rare codons has been recognized for a long time, if a particular domain in a protein needs extra time to fold, then sometimes a rare codon will be selected for downstream in the transcript, to make the ribosome pause for a little while 09:28 < kanzure> "DNA punch cards for storing data on native DNA sequences via enzymatic nicking" https://www.nature.com/articles/s41467-020-15588-z 09:29 < kanzure> https://www.scientificamerican.com/article/punch-card-dna-could-mean-cheaper-high-capacity-data-storage/ 10:05 < kanzure> https://joinmarket.me/blog/blog/schnorrless-scriptless-scripts/ 10:06 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 265 seconds] 10:24 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 10:31 < kanzure> jrayhawk: fenn: i think this is the subset for redirects? https://diyhpl.us/~bryan/irc/bitcoin/bitcoin-transcripts.txt 10:57 < jrayhawk> Separating out the btc transcripts will mean less discoverability on the diyhplus stuff, if that matters to you. SEO optimization seems unnecessary for lack of competition. Topical aggregation could be done with ikiwiki tags and native rss generation instead. 11:12 < jrayhawk> https://ikiwiki.info/ikiwiki/directive/inline/ https://ikiwiki.info/ikiwiki/directive/tag/ 11:12 < jrayhawk> I'm not advocating for anything in particular, here, just pointing out options. 11:33 < kanzure> i think he wants to make a prettylooking website and i wouldn't blame him if he doesn't want to figure out ikiwiki css 11:33 < kanzure> git subtree might be appropriate 11:34 < kanzure> no harm with trying inline 11:53 < jrayhawk> oh, he's actually setting up an SEO-competing site 12:43 -!- Human_G33k [~HumanG33k@62.147.242.8] has quit [Remote host closed the connection] 12:47 -!- filipepe [uid362247@gateway/web/irccloud.com/x-zzucupwdlhxdydby] has joined ##hplusroadmap 12:48 -!- HumanG33k [~HumanG33k@62.147.242.8] has joined ##hplusroadmap 12:49 < kanzure> i guess 12:49 < kanzure> soo it's not like anyone can remember how to spell diyhpl.us anyway 12:49 < kanzure> and most people can't seem to navigate it 12:49 < kanzure> and why is it hanging off of our wiki anyway? 12:50 -!- HumanG33k [~HumanG33k@62.147.242.8] has quit [Remote host closed the connection] 12:50 -!- HumanG33k [~HumanG33k@62.147.242.8] has joined ##hplusroadmap 12:53 < kanzure> .tw https://twitter.com/MartinAndler/status/1249021279660683265 12:53 < saxo> Very sad to have learned the passing away today of the famous mathematician John Conway from Covid-19. (@MartinAndler) 13:36 -!- yashgaroth [~ffffffff@172.58.19.12] has quit [Ping timeout: 240 seconds] 13:52 -!- yashgaroth [~ffffffff@172.58.19.12] has joined ##hplusroadmap 14:34 < kanzure> "Attacking with bitcoin: Using bitcoin to build resilient botnet armies" https://arxiv.org/pdf/2004.01855.pdf 15:47 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 256 seconds] 15:54 -!- darsie [~kvirc@84-114-73-160.cable.dynamic.surfer.at] has quit [Ping timeout: 260 seconds] 15:59 -!- superkuh [~superkuh@unaffiliated/superkuh] has quit [Remote host closed the connection] 16:00 -!- superkuh [~superkuh@c-24-118-172-137.hsd1.wi.comcast.net] has joined ##hplusroadmap 16:00 -!- superkuh [~superkuh@c-24-118-172-137.hsd1.wi.comcast.net] has quit [Changing host] 16:00 -!- superkuh [~superkuh@unaffiliated/superkuh] has joined ##hplusroadmap 16:01 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 16:20 -!- balrog [~balrog@unaffiliated/balrog] has quit [Quit: Bye] 16:26 -!- balrog [~balrog@unaffiliated/balrog] has joined ##hplusroadmap 18:33 -!- CryptoDavid [uid14990@gateway/web/irccloud.com/x-qkjmdpkxkfmikcgk] has quit [Quit: Connection closed for inactivity] 18:59 -!- dongcarl [~dongcarl@unaffiliated/dongcarl] has joined ##hplusroadmap 18:59 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Quit: leaving] 19:02 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Remote host closed the connection] 19:16 -!- filipepe [uid362247@gateway/web/irccloud.com/x-zzucupwdlhxdydby] has quit [Quit: Connection closed for inactivity] 20:18 -!- yashgaroth [~ffffffff@172.58.19.12] has quit [Quit: Leaving] 21:44 -!- pixeldust [~pixeldust@50-39-183-114.bvtn.or.frontiernet.net] has joined ##hplusroadmap 23:03 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 23:13 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Quit: leaving] --- Log closed Sun Apr 12 00:00:49 2020