--- Day changed Tue Jun 03 2008 00:42 < joshcryer> !help 00:42 < joshcryer> Denied! :( 00:44 < kanzure> joshcryer: Hello. 00:48 < Vedestin> krebs is not servicing the channel it seems 00:49 < kanzure> do we want krebs servicing the channel? 00:50 < Vedestin> did it do anything 00:51 < kanzure> it told epitron secrets. 00:51 < Vedestin> hmm 00:52 < joshcryer> Hey kanzure 00:53 < kanzure> What's up? 00:53 < joshcryer> Listening to the Phoenix briefing cast got me thinking. 00:53 < joshcryer> They talked about TEGA and how it seperated partciles by ionizing them and using magnetic fields to push them along. 00:54 < joshcryer> (Since matter itself is non-magnetic without being ionized.) 00:54 < joshcryer> (briefing cast = the mp3 of the phone briefing they had) 00:54 < joshcryer> Anyway, it sounded interesting enough to discuss. 00:54 < kanzure> I don't understand. The magnetic fields would push the ions along? 00:55 < joshcryer> Yeah 00:55 < joshcryer> They heat up some soil. 00:55 < kanzure> I think the idea of the ion Hull reactors used to be that the ions would eventually just squeeze their way out or something. 00:55 < joshcryer> It breaks down into the constituant compounds (it's not hot enough to make them split up into their elements). 00:55 < kanzure> Oh, the actual lander. 00:55 < joshcryer> And since a water particle isn't exactly magnetic they ionize it. 00:56 < joshcryer> So that magnetic fields can interact and they can hold it in a nice plasma for spectromoy. 00:56 < kanzure> Yes, yes, but once they have that plasma, how is this used for propulsion ? 00:56 < kanzure> I mean, pushing the ions along, this would create thrust, yes? 00:57 < joshcryer> We're not talking about propulsion. :P 00:57 < joshcryer> We're talking about soil analysis. :) 00:57 < kanzure> bleh 00:57 < kanzure> Okay :) 00:57 < joshcryer> I think it applies though. 00:57 < joshcryer> We know the carbothermic reaction can seperate elements pretty effectively. 00:57 < kanzure> So, mass spec with magnets? 00:57 < joshcryer> The question is, how do you engineer something that does the chemistry? 00:57 < kanzure> or at least magnets to do separation, hrm 00:57 < joshcryer> That's what I was thinking. 00:57 < kanzure> What's the exact carbothermic reaction? 00:58 < joshcryer> Magnetic centrafuge thing. 00:58 < joshcryer> It's on page 9 of that paper. 00:58 * kanzure is ashamed that he hasn't gotten to it. 00:59 < joshcryer> Pfft, it's a nice diagram. 00:59 < kanzure> Actually, my time has been spent reviewing the classic: 00:59 < kanzure> http://heybryan.org/~bbishop/docs/ellingtonia/2008-03-23/Construction%20of%20an%20in%20vitro%20bistable%20circuit%20from%20synthetic%20transcriptional%20switches.pdf 00:59 < kanzure> It's for a group I just joined today -- first day at the lab. yay \o/ 01:00 < kanzure> http://en.wikipedia.org/wiki/Carbothermic_reaction 'Carbothermic reactions are thermic chemical reactions which use carbon as the reducing agent at high temperature. The most prominent example is used in iron ore smelting.' 01:00 < kanzure> hrm. 01:02 < joshcryer> To me the paper itself is pretty much irrelevant. 01:02 < joshcryer> It just proves what we already know. 01:02 < joshcryer> http://i32.tinypic.com/14ufaes.png 01:02 < joshcryer> What I think is important, at least to me, perhaps it's stupid, is the element seperation cycle expounded upon there. 01:03 < kanzure> Uhh, where'd that come from? 01:03 < kanzure> I mean, who acquired that information and how ? 01:03 < joshcryer> It's from the paper I sent you... page 9. 01:03 < kanzure> right, but 01:03 < joshcryer> Lackner wrote the paper, but I'd have to look at it more closely to see where he got the chemistry. 01:04 * kanzure grumbles. Guess I really have to read it now. 01:04 < joshcryer> JANAF thermodynamic tables. 01:05 < joshcryer> http://www.nist.gov/srd/jpcrd_28.htm 01:05 < joshcryer> Holy shit. 01:05 < joshcryer> The book is $305. 01:05 < kanzure> heh' libraries at universities will have it. 01:06 < joshcryer> To check out? 01:07 < Vedestin> shortloans atleast 01:07 < joshcryer> Probably not for non-students though, I imagine? 01:07 < joshcryer> (Obviously I'd want to scan it in. ;)) 01:07 < Vedestin> not to check out, now 01:07 < Vedestin> they have scanners in uni libraries 01:07 < joshcryer> Barns and Noble has it for $195. 01:07 < joshcryer> I'll see if the local universities have it. 01:08 < Vedestin> yeah, university has it for free 01:08 < kanzure> joshcryer: I've been able to walk into uni libraries quite easily. 01:08 < kanzure> portable scanner :) 01:08 < joshcryer> kanzure, same here. 01:09 < Vedestin> anyone can walk in to a uni library 01:09 < joshcryer> All schools here are open at least. 01:09 < joshcryer> Vedestin, well some schools are crap and locked down because of the shootings. 01:09 < Vedestin> isnt that just high schools 01:09 < Vedestin> that's barely a uni at all 01:09 < joshcryer> V. Tech is all locked down now. 01:09 < Vedestin> who the fuck would lockdown 01:10 < kanzure> uh 01:10 < kanzure> virginia tech got slaughtered 01:10 < joshcryer> Yeah 01:10 < Vedestin> i remember that 01:10 < Vedestin> cho 01:11 < joshcryer> Worst in American history, so it's no surprise they locked dow. 01:11 < Vedestin> cho was a student 01:11 < Vedestin> he'd have gotten in anyway 01:11 < joshcryer> But yeah when I had no computer I would walk into uni library and just use internet. 01:11 < joshcryer> They don't bug you unless it's exams time. 01:11 < joshcryer> Then they'll ask if you're a student. 01:11 < Vedestin> are you homeless? 01:11 < joshcryer> LOL, no. 01:12 < Vedestin> why would they bug you at all then 01:12 < joshcryer> I went on, a waltzing matilda before though. 01:12 < Vedestin> i see 01:12 < joshcryer> Because if it's exams time and you look like a non-student and computers are needed they'll ask you for your student ID. 01:12 < Vedestin> libraries are always full of mature-age students 01:13 < Vedestin> because kids have their own computers 01:13 < kanzure> uh? 01:13 < joshcryer> Yeah 01:13 < joshcryer> Now adays it doesn't really matter. 01:13 < joshcryer> Students use their own computers. 01:13 < kanzure> Have you been to a public high school library ? 01:13 < joshcryer> But back in 2002 I went to a university and it was exams time and they booted me. 01:14 < Vedestin> public high school isn't university 01:14 < joshcryer> It was in Clearwater OK. 01:14 < joshcryer> Can't remember the university. 01:14 < Vedestin> lol 01:14 < joshcryer> But w/e, it's a moot point. 01:14 < joshcryer> I'll see if The University of Colorado has a copy. 01:15 < joshcryer> Let's see if I can get on their website. 01:15 < joshcryer> Barns & Nobel has it for $160. 01:15 < Vedestin> buy it if you can afford it 01:16 < joshcryer> I think I might. 01:17 < Vedestin> i just failed first semester 01:17 < joshcryer> Aww. 01:17 < joshcryer> What were you taking? 01:18 < Vedestin> math 111, chem 101, engineering 100 and phys 120 01:18 < Vedestin> i withdrew from phys straight up 01:18 < Vedestin> then i withdrew from eng 01:18 < joshcryer> You exploded bro. 01:18 < Vedestin> now i realise i have my final exam for math next week and i haven't been to a lecture in 3 months 01:18 < joshcryer> Gotta take it easy. 01:18 < joshcryer> That's some hard stuff. 01:18 < Vedestin> four courses per semester is the norm 01:19 < Vedestin> plus, i've been out of high school for a few years, so i couldn't remember anything 01:19 < Vedestin> only thing i'll pass is chem 01:20 < joshcryer> Library use only. 01:20 < joshcryer> That's a shame. 01:20 < Vedestin> there are scanners in the library 01:20 < Vedestin> flatbeds atleast 01:20 < Vedestin> or there should be 01:20 < joshcryer> The 1986 version is for checkout though. 01:20 < joshcryer> I imagine that one is good enough. 01:20 < joshcryer> Who needs the 1998 version. 01:21 < Vedestin> get that then 01:22 < joshcryer> I need to find someone enrolled... ;) 01:25 < Vedestin> oh yeah, right 01:26 < Vedestin> you could just steal a wallet 01:26 < joshcryer> Nah. 01:26 < Vedestin> i bet it's a self checkout service 01:26 < joshcryer> I'm sure someone would be helpful. 01:26 < Vedestin> really? 01:26 < joshcryer> In fact my brother is enrolling there this summer. 01:26 < Vedestin> there's no way i'd let some random check a book out on my record 01:26 < joshcryer> I could just wait. 01:26 < joshcryer> Or buy it at B&N. 01:26 < joshcryer> Vedestin, even if I offered you dinner and let you watch me scan the tables in? 01:27 < joshcryer> (Not the whole book, just the important parts.) 01:27 < Vedestin> would you cook yourself? 01:27 < joshcryer> Haha 01:27 < joshcryer> Why yes, yes I would. 01:27 < joshcryer> What would you like to eat? 01:27 < Vedestin> nah, i don't want you to cook 01:27 < Vedestin> we'll just get chinese 01:27 < Vedestin> seriously though, there'd be scanners at the library 01:28 < joshcryer> Yeah I know. 01:28 < joshcryer> I'll go check it out. 01:28 < Vedestin> k bye 01:28 < joshcryer> Haha 01:28 < joshcryer> It's 11:30PM here. 01:28 < Vedestin> uni libraries are 24 hour 01:29 < joshcryer> Yeah but. 01:29 < joshcryer> I think I'd be more obvious going in at night. 01:33 < Vedestin> maybe they have a security dude or something 01:34 < Vedestin> couldnt you just get a library card from them 01:34 < joshcryer> Haha 01:35 < Vedestin> not from the security guard 01:35 < Vedestin> from the university 01:35 < Vedestin> enrol in a non-award program 01:35 < Vedestin> get a student card 01:35 < Vedestin> and you're done 01:35 < Vedestin> free books 01:35 < Vedestin> forevers 02:07 < joshcryer> http://www.nist.gov/srd/PDFfiles/jpcrdM9.pdf 02:07 < joshcryer> It's online! 02:07 < joshcryer> How did I overlook that? 02:07 < joshcryer> 250mb pdf file lol 02:18 < Vedestin> hahaha 02:20 < joshcryer> if this was OCR'd it'd be 25 mb 02:21 < ybit> http://books.google.com/books?id=8axpjcE0bqUC&dq=%E2%80%9CBrain-Implantable+Biomimetic+Electronics+as+a+Neural+Prosthesis+for+Hippocampal+Memory+Function,%E2%80%9D&source=gbs_summary_s&cad=0 02:22 < ybit> 1:22am here 02:22 < ybit> thought i'd chime in :) 02:49 < joshcryer> This book has every single chemical composition known to man. 02:49 < joshcryer> And formulas how to get to each one. 02:49 < joshcryer> This is all you need! 03:59 < joshcryer> Any other channels like this that aren't so dead? 05:33 -!- nsh [n=nsh@eduroam-80.uta.fi] has joined #hplusroadmap 06:17 -!- joshcryer [n=g@unaffiliated/joshcryer] has quit [Remote closed the connection] 07:31 -!- faceface [n=chatzill@bioinformatics.org] has joined #hplusroadmap 07:32 < faceface> can you give me an implant to speak german? 07:36 < nsh> the you that speaks german is discontinuous from the you that doesn't 07:44 -!- fenn_ [n=pz@adsl-76-251-87-73.dsl.bltnin.sbcglobal.net] has joined #hplusroadmap 07:44 -!- Topic for #hplusroadmap: http://heybryan.org/ http://heybryan.org/mediawiki/ http://heybryan.org/exp.html | krebs is now servicing the channel. try !help 07:44 -!- Topic set by kanzure [] [Tue Apr 29 18:54:31 2008] 07:44 [Users #hplusroadmap] 07:44 [ faceface] [ fenn_ ] [ nsh ] [ Vedestin] 07:44 [ fenn ] [ kanzure] [ Phreedom] [ ybit ] 07:44 -!- Irssi: #hplusroadmap: Total of 8 nicks [0 ops, 0 halfops, 0 voices, 8 normal] 07:44 -!- [freenode-info] if you need to send private messages, please register: http://freenode.net/faq.shtml#privmsg 07:45 -!- Channel #hplusroadmap created Sat Mar 22 15:44:12 2008 07:45 -!- Irssi: Join to #hplusroadmap was synced in 37 secs 08:01 -!- fenn [n=pz@adsl-76-251-84-248.dsl.bltnin.sbcglobal.net] has quit [Read error: 110 (Connection timed out)] 08:13 < kanzure> faceface: Hey. 08:20 < faceface> yo k 08:20 < faceface> I figure, what would I really want an implant for? 08:20 < faceface> I would like a dictionary of german words and phrases 08:21 < faceface> how could we hook up such a memory expansion? 08:21 < faceface> would that 'brain interferance' technology that got posted a while back (the one that stopped a guy from talking using a targeted magnetic field) be used in a refined way to 'implant' certain thoughts? 08:22 < faceface> let me put my thinking cap on... 08:22 < Vedestin> it wouldn't 08:22 < faceface> Vedestin: wouldn't what? 08:22 < Vedestin> it's like how you use the great wall of china to do brain surgery 08:22 < faceface> Vedestin: sure, at current resolutions 08:23 < faceface> Won't the resolution get driven down by better magnets (super conductors) and better software 08:23 < faceface> ? 08:23 < Vedestin> probably not 08:23 < Vedestin> not to a level that you'd need atleast 08:23 < faceface> MRI -> MFRNI 08:24 < faceface> Vedestin: but what 'level' would be interesting? currently you can target the 'speach centers', surly we can do more intereseting thigs with higher res 08:25 < faceface> what about focusing attention on the 'forign language' centers of the brain? (I heard that adults who become bilingual use a different region of their brain to talk forign than they do to talk their mother tongue) 08:26 < faceface> what about mildly surpressing the 'mother tongue' brain function while mildly stimulating the forign language centers during an intensive dictionary crash course? 08:26 < faceface> of course it could backfire and leave you talking less of the forign language than controll 08:26 < faceface> Vedestin: my point is that refining the 'level' will always be 'interesting', therefore high levels of resolution should be acheived naturally 08:27 < faceface> by continuious improvement driven by better technology 08:27 < faceface> anyway... let me know when we can start de deutsche sprachen. 08:32 -!- You're now known as fenn 08:32 < Vedestin> faceface, it really doesn't sound like the way to go 08:36 < kanzure> http://en.wikipedia.org/wiki/Friendly_artificial_intelligence <-- I see. Eli got the nod from a few 'professionals'. 08:36 < kanzure> faceface: http://heybryan.org/mediawiki/index.php/rTMS 08:36 < kanzure> faceface: Your idea of supressing the mother tongue is interesting. 08:37 * nsh is dubious 08:37 < kanzure> nsh: It would be fun to try. 08:37 < nsh> just wondering about children who are raised bilingually 08:38 < nsh> also, don't you mean: it'd be fun to try on a neural slice or something else that makes your earlier disparagement of chemical neuromodification experiments non-hypocritical? :-) 08:43 < kanzure> http://www.singinst.org/ourresearch/publications/what-is-friendly-ai.html 08:43 < kanzure> nsh: No, I really mean it'd be fun to use rTMS :) 08:45 < fenn> faceface: the problem with rTMS is it's like using a stick of dynamite, when what you really want is a chisel 08:45 < fenn> and there's no obvious way to make the effect more fine grained 09:00 < kanzure> I want a proof about this 'friendly ai' BS. I think the unstated fear is "well, Vinge said they will be able to catch us no matter where we go, so we're doomed!" 09:01 < kanzure> doomed 10:12 -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit ["Leaving."] 10:15 -!- Vedestin [n=Vedestin@d122-109-35-58.sbr3.nsw.optusnet.com.au] has quit [Read error: 104 (Connection reset by peer)] 10:36 < faceface> fenn: how was it 'focused' in the first place? 10:36 < faceface> nsh: bilingual individuals use the same region of their brain for both languages 10:36 < faceface> (unlike ppl. who learn a language in adulthood) 10:37 < faceface> perhaps we could just develope a 'learn a language pill' that lets you access the 'core' language region of your brain again... or perhaps that would just result in turning you into a babbeling child 10:51 < nsh> so.. beyond which stage of cognitive development does a second language require an alternate localisation? 10:52 < nsh> say if you were raised bilingually until the age of 7, then spoke just one language until adolescence, then relearnt the second language... 10:52 < nsh> (as is my case) 10:52 * nsh remains dubious 10:57 < faceface> nsh: interesting example 10:57 < faceface> you would need to be probed ;-) 10:58 < faceface> I only remember the factoid 'ppl. use a different region of their brain to talk a second language if learnt after childhood' 10:58 < faceface> I could guess 7 years... but its a guess 10:59 < faceface> in your case... I'd say that the bilingual impression for 7 years would not fade 10:59 < faceface> you went back to your core region... I would guess.... now if you learned a 3rd language... oh my! 11:01 < nsh> thing is that these summarisations of research are generally more misleading than revealing 11:01 < faceface> how dare you! my left brain is insulted! 11:02 < faceface> my right brain is admiring my desktop wallpaper 11:02 < faceface> http://www.wellcome.ac.uk/News/News-archive/Browse-by-date/2007/News/WTD026366.htm ?? 11:02 * nsh smiles 11:06 < faceface> It has been shown that monolingual children lose at least some of their sensitivity to foreign phonemes between the age of 6 and 12 months (Werker and Tees 1984Go; Cheour and others 1998Go; Rivera-Gaxiola and others 2005Go). ??? 11:06 < faceface> Pallier and others (1997)Go have shown that even early exposure to a second language before the age of 6 years is not enough to attain first language–type phonological competence. 11:08 < faceface> http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6WC0-49JHHF5-6&_user=28761&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000002718&_version=1&_urlVersion=0&_userid=28761&md5=7ea6388b87abe31b66062c2cf6fe7c57 11:08 < faceface> that last one is as close as I have found --- Log opened Tue Jun 03 11:22:22 2008 11:23 -!- fenn [n=pz@adsl-76-251-81-170.dsl.bltnin.sbcglobal.net] has joined #hplusroadmap 11:23 -!- Topic for #hplusroadmap: http://heybryan.org/ http://heybryan.org/mediawiki/ http://heybryan.org/exp.html | krebs is now servicing the channel. try !help 11:23 -!- Topic set by kanzure [] [Tue Apr 29 18:54:31 2008] 11:23 [Users #hplusroadmap] 11:23 [ faceface] [ fenn] [ nsh] [ Phreedom] [ ybit] 11:23 -!- Irssi: #hplusroadmap: Total of 5 nicks [0 ops, 0 halfops, 0 voices, 5 normal] 11:23 -!- Channel #hplusroadmap created Sat Mar 22 15:44:12 2008 11:23 -!- Irssi: Join to #hplusroadmap was synced in 32 secs 11:29 < faceface> fenn: use less dynamyte 11:29 < faceface> fenn: or build something else 11:39 < fenn> the nice thing about rTMS is you dont have to drill any holes in peoples' skulls 11:41 < fenn> i bet there's some way to use the bessel function to recreate a desired 3D electric field in arbitrary space (but only for a short amount of time, which might still work) 11:42 < fenn> or whatever the 3d version of the bessel function is called 11:44 -!- nsh is now known as manglefango 11:44 -!- manglefango is now known as nsh 11:52 < fenn> this is what i'm remembering re: bessel function: http://www.pinktentacle.com/2006/07/device-uses-waves-to-print-on-water-surface/ 11:52 < nsh> that's cool 11:53 < nsh> could be cooler 11:54 < fenn> aww the video's down wtf 11:54 < fenn> youtube gets suckier every day 11:54 < nsh> if i was allowed anywhere near google 11:55 < nsh> removed videos would automatically redirect to replacements 11:56 < nsh> then again, i would probably not have videos go down, ever 11:56 < nsh> but i think people would dislike that even more 11:57 < fenn> people with lawyers. i think regular people wouldn't care 11:58 < fenn> that think is sorta like a phased array radar 12:00 < nsh> hm 12:03 < nsh> bessel functions also arise in quantum mechanics 12:03 < nsh> http://quantummechanics.ucsd.edu/ph130a/130_notes/node225.html 13:01 -!- faceface [n=chatzill@bioinformatics.org] has quit [Nick collision from services.] 13:02 -!- facefaceface [n=chatzill@bioinformatics.org] has joined #hplusroadmap 13:33 -!- joshcryer [n=g@unaffiliated/joshcryer] has joined #hplusroadmap 14:21 -!- ybit [n=u1@unaffiliated/ybit] has quit [Connection timed out] 15:26 -!- Splicer [n=p@h34n1c1o261.bredband.skanova.com] has joined #hplusroadmap 15:33 -!- ybit [n=u1@unaffiliated/ybit] has joined #hplusroadmap 15:37 -!- nsh [n=nsh@wikipedia/nsh] has quit [Success] 15:46 -!- nsh [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has joined #hplusroadmap 15:50 -!- kanzure [i=bryan@dhcp-146-6-213-183.icmb.utexas.edu] has joined #hplusroadmap 15:50 -!- kanzure [i=bryan@dhcp-146-6-213-183.icmb.utexas.edu] has quit [Read error: 104 (Connection reset by peer)] 18:06 -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap 18:07 < kanzure> Hi all. 18:07 < fenn> g'day 18:39 < kanzure> http://grailsearch.org/?q=node/77#comment-226 <-- I just posted to grailsearch.org 18:40 < kanzure> fenn: http://heybryan.org/mediawiki/index.php/Wittig_cohort but probably stuff you've already seen 18:40 < kanzure> I'm pretty much very, very sure now that the majority of good work that happens, happens on your own 18:41 < kanzure> since the majority of the work that I got done in the day was in a 5 minute period. 18:42 < fenn> 5 minutes in the lab? 18:43 < kanzure> well, on a computer 18:44 < kanzure> I'm not doing anything wet since they want to do some "Don't Die" safety training courses 18:44 < fenn> oh, dangerous stuff, that DNA 18:44 < kanzure> which is totally BS, and they know it, they're actually waiting for a student to come back in who knows his techniques/protocols well 18:44 < kanzure> right, we're not even doing radiotagging 18:46 < kanzure> bbl 18:52 < kanzure> okay, back 18:52 < kanzure> fenn: I'm thinking of illustrating skdb with wet ware first, even though wetware has this tendency to suck 18:52 < kanzure> and generally not be solid state 18:52 < kanzure> so the 'bioreactor'/biotank thing 18:53 < fenn> looks like a pain in the butt 18:53 < fenn> oops 18:53 < kanzure> yes 18:53 < kanzure> but 18:53 < fenn> wrong chan sry 18:53 < kanzure> it'll have immediate results in that it'll be mostly self-replicating 18:53 < kanzure> except for the metal tank 18:53 < kanzure> and uh, maybe some other things like any electronics (which I want to keep to a minimum -- have it all automated via biology) 18:54 < kanzure> actually, the tank may not be metal 18:54 < kanzure> but you get the idea 18:55 < fenn> hmm.. automated via biology? 18:55 < kanzure> so, the baseline system will just produce the chemicals like T7 and so on, that would be necessary for transcription and doing other DNA experiments, gene splicing, insertion, plasmids, whatever 18:56 < kanzure> you were mentioning in vitro synthesis via laser light flashing a few months ago -- I think I've figured out some possible implementations 18:56 < kanzure> the next step up from baseline would do something, I don't know what. Personally I want to try inserting the neurotransmitter synthesis pathways into ecoli. 18:57 < fenn> kainic acid stuff? i think if it were easy someone would have done it already, but maybe its just cross-discipline taboo stuff 18:57 < kanzure> right, I'm not completely certain it's possible, I bet there's a lot of funky missing pieces 18:57 < kanzure> the purification proc migth be it 19:13 < kanzure> fenn: looks like I might have a contract with somebody to extend drupal.org software 19:14 < kanzure> hurray for PHP? 19:14 < fenn> a contract? 19:14 < kanzure> well, I'm waiting for him to email it to me 19:14 < kanzure> but I just got off the phone and it sounds good 19:16 < kanzure> uhm, so, the purification stuff is what I'm worrying about 19:16 < kanzure> affinity chromatography sounds like it takes some pretty funky equipment 19:18 < fenn> its not just a column with various solutions passing through it? 19:18 < kanzure> ok, please explain column chromatography setups 19:18 < kanzure> I've read wikipedia, I've read other docs, it just doesn't make sense 19:18 < kanzure> paper chromatography, being the first type I learned about, has completely thrown me off base 19:19 < fenn> so there's a column, meaning a cylindrical plug of goo in a syringe or a glass tube 19:20 < fenn> you can add water to the column and it slowly oozes through the plug of goo 19:20 < fenn> the plug is the 'substrate' which has an affinity for what you're trying to purify 19:20 < kanzure> goo could equal wha? something like, a gel? 19:20 < fenn> so, you'd bind your DNA to that 19:20 < kanzure> hrm 19:20 < fenn> yeah it looks like agarose (at least when i did it) 19:21 < kanzure> this is column? 19:21 < fenn> yes 19:21 < kanzure> okay, so how do we come up with affinities for specific molecules? 19:21 < kanzure> computational chem? 19:21 < fenn> natural selection? 19:21 < fenn> aptamer process 19:22 < fenn> mm.. i guess you have to have a sample of what you want before you can do that 19:22 < kanzure> aptamer process sucks apparently, I've asked the Ellington goons 19:22 < kanzure> I mean, for purification 19:22 < fenn> why? 19:22 < kanzure> for selection, like bonding to receptors, it works 19:23 < kanzure> same reason you mentioned 19:23 < fenn> not specific enough? 19:23 < kanzure> 'sample of what you want before in the first place' 19:23 < kanzure> specificity has to increase with bp length :( 19:23 < kanzure> so 100 bp might not be enough in some cases, apparently 19:23 < fenn> well once you have the aptamer then you sequence it and anyone can use it 19:23 < kanzure> and as you increase bp, the DNA synthesizers quickly begin to suck 19:23 < kanzure> (in vitro DNA synthesis, again, might be useful) 19:23 < kanzure> right, or just copy them 19:24 < kanzure> aptamer -> denature -> clone -> ... 19:24 < fenn> can use a western blot to purify small samples of protein 19:24 < fenn> works sorta like 2 dimensional paper chromatography 19:24 < fenn> you end up with a piece of paper with dots all over it 19:24 < kanzure> I've heard about blots, but I still have't internalized 19:24 < kanzure> oh, blot == just a gel? 19:24 < fenn> its filter paper 19:24 < fenn> but works sorta like a gel 19:25 < kanzure> so the filter is specific to what you're blotting? 19:26 < fenn> oh sorry the actual separation is performed in a gel 19:26 < kanzure> oh 19:26 < fenn> and then it gets 'blotted' onto filter paper 19:26 < kanzure> wait, I misinterpreted your recommendation 19:26 < kanzure> "natural selection? aptamer process" 19:26 < kanzure> I didn't read the implied "i.e., aptamer process" 19:26 < kanzure> the implied "i.e." 19:26 < kanzure> hrm, the same comments we mentioned still apply :) nevermind 19:29 < kanzure> so, bruteforcing purification methodologies: physical/mechanical, chem, electrical, thermo, nuclear (radioactivity & decay), surface tension (according to Wikipedia article on energy) 19:29 < fenn> since its a recombinant protein, you could 'tag' the protein with a covalently linked 'handle' of a well-characterized binding complex 19:30 < kanzure> physical - crank up the RPMs via centrifuges, phase changes (affinity chromatography), 19:30 < kanzure> well, it's multiple proteins of course 19:30 < fenn> and then just cleave the handle off with a digestion enzyme 19:30 < fenn> you could use the same method each time 19:30 < kanzure> oh 19:30 < kanzure> what if the handle is degradable? 19:30 < kanzure> that would be useful 19:30 < kanzure> once you have it purified, you degrade the handle, inject into human 19:30 < kanzure> yeah, cleave 19:31 < kanzure> so, there's a known binding complex that we can use? 19:31 < fenn> Adding a tag to the protein such as RuBPS gives the protein a binding affinity it would not otherwise have. Usually the recombinant protein is the only protein in the mixture with this affinity, which aids in separation. The most common tag is the Histidine-tag (His-tag), that has affinity towards nickel or cobalt ions. Thus by immobilizing nickel or cobalt ions on a resin, an affinity support that specifically binds to histidine-tagged proteins can be created. 19:32 < kanzure> cutoff@histidin (support that specifically binds to histidin) 19:32 < fenn> right 19:32 < kanzure> but I get it 19:32 < fenn> there is probably a more biological version that works similarly 19:32 < kanzure> let's say that we have a tank of cells (or even just a tank of ribo* stuff, but I think we need cells still) that have transcribed & translated & made the protein we want, with the RuBPS 19:33 < kanzure> so, how would we do this? would we cycle water over the culture, 19:33 < kanzure> or would we centrifuge the hell out of it? 19:33 < fenn> i like those beads actually 19:33 < kanzure> and then take the remaining supernatant / goo and run it through something that has surface bonds of 19:33 < kanzure> yeah, so beads on a surface? 19:33 < kanzure> actually, I suggested earlier today to a friend to use nanoparticles in the blood stream 19:33 < kanzure> for something totally different 19:33 < fenn> beads in solution, centrifuge, decant, elute beads 19:33 < kanzure> if the nanoparticles could be used to bind to the RuBPS handles 19:34 < kanzure> in context, we were discussing myostatin and muscle growth in marathon mice and hu 19:34 < kanzure> the poor grad student (Mike Wittig) has realized that I'm pretty well read 19:34 < fenn> heh 19:34 < kanzure> fun guy, Mike. 19:35 < kanzure> btw, Mike has a friend at UT somewhere, some Russian guy named Stan, that does multimonitor setups. he also did an IP-by-IP scan of the women's dorms to figure out which rooms had linux installations ... 19:35 < kanzure> heh' 19:36 < kanzure> IP blocks rock in that way. 19:36 < fenn> dunno what to think about that 19:36 < kanzure> fenn: don't you and I googlestalk anyway? 19:36 < fenn> i plead the fifth 19:36 < kanzure> anyway 19:36 < kanzure> how do you get the molecules off of the beads 19:37 < fenn> cleave the tag 19:37 < kanzure> the tag is in between the molecule and the surface 19:37 < fenn> or some salty magick 19:37 < kanzure> not physically accessible 19:37 < fenn> i was thinking there's a long chain between the two proteins 19:37 < kanzure> hrm, salty magik -> degrade the proteins, they detach, but then reform when solution returns to normal ? 19:37 < kanzure> perhaps 19:37 < kanzure> I'm not sure, how would we make sure? is that common? 19:37 < fenn> make sure what? 19:37 < kanzure> that there's a good space between the protein and the nanoparticle 19:37 < kanzure> hrm, buckyballs 19:38 < fenn> uh.. you write it into the sequence 19:38 < kanzure> huh? 19:38 < kanzure> the protein can just randomly be mutated like that? 19:38 < fenn> gene start... protein A... chain ... protein B ... end 19:38 < kanzure> to be given such a 'tail' ? 19:38 < kanzure> cool 19:38 < kanzure> I think that's something involving hydrogen bonds 19:38 < kanzure> I remember something about specific protein bonding mechanisms 19:39 < fenn> you could build it up that way, but it's a lot more work 19:42 < kanzure> the beads are ridiculously easy to synthesize 19:43 < kanzure> now, the surface chem to bind the RuBPS-gropers to the nanoparticle? 19:43 < kanzure> actually, it doesn't need to be nano 19:43 < kanzure> but nano maximizes surface area 19:45 < kanzure> http://postbiota.org/pipermail/tt/2008-February/002547.html only thing wrong with this page - Gold is not a semiconductor. 19:45 < kanzure> I quickly corrected myself: http://postbiota.org/pipermail/tt/2008-March/002548.html 19:52 < kanzure> so, fenn, remember how you were talking about in vivo sequencing/synthesis a while back? 19:52 < Splicer> i know very little about protein purification... i just started reading about it a few days ago... my image of it was that the standard method was affinity binding with receptor coated plastic beans in columns.. and that the receptors were synthezized to match the protein one wanted bound... am I totally off? Or is there some reason the 'add standard tag - catch the proteins with it - cleave the tag off' procedure is not used? 19:52 < kanzure> fenn: what we're building in the lab is a way to translate frequency (like flashing photon input) into kinetical changes; 19:53 < kanzure> perhaps this could be a way to control transcription? 19:53 < kanzure> I don't actually recall the exact method that you were suggesting 19:56 < fenn> sure, hamad schifferle (?) at CBA was doing gold nanoparticles that caused DNA to unravel when exposed to radio 19:56 < fenn> god CBA's website really sucks 19:57 < fenn> unless that was some weird dream i'm remembering 19:57 < kanzure> no, CBA really sucks 19:57 < kanzure> (web) 19:58 < fenn> Remote Electronic Control of DNA Hybridization Through Inductive Coupling to an Attached Metal Nanocrystal Antenna, K. HamadSchifferli, 19:59 < fenn> ah its one of those hybrid frankenstein names 20:00 < kanzure> okay, that's close 20:00 < fenn> http://web.mit.edu/be/people/hamad.htm 20:00 < kanzure> but that's requiring metal 20:00 < kanzure> we're building a biological equivalent 20:00 < fenn> yeah should be able to do similar with quinones, fluorescins, rhodopsin, chlorophores 20:00 < fenn> basically anything colorful 20:01 < fenn> maybe could hack some existing light-activated enzymes 20:02 < kanzure> nono, 20:02 < kanzure> we have a represillator/oscillator -- a circuit of transcriptional switches going around in a circle 20:02 < kanzure> light-activated enzymes would be useful, I suppose 20:02 < fenn> i dont know what 'represillator' means 20:02 < kanzure> but it's not in the transcriptional logic itself 20:02 < kanzure> represillator == oscillator 20:02 < kanzure> represillator = repressor/activator oscillator 20:02 < fenn> ok 20:02 < fenn> stupid name 20:02 < fenn> ;) 20:02 < kanzure> heh :) 20:03 < fenn> so, that's your clock cycle? 20:03 < kanzure> yes 20:03 < kanzure> fenn: I could send you a few papers 20:04 < fenn> then you only need one 'data line' for asynchronous serial 20:04 < kanzure> I'll have to show you the paper, yeah 20:04 < fenn> i think the hard part is getting it all to work :) 20:05 < kanzure> sure 20:05 < kanzure> we're doing just one section of the larger schematic at the top of the page on, uh, page 7 20:05 < kanzure> but I'm dubious about the memory state system -- I think it can work, but it's hard to get people understanding that it's not quite like physical RAM address space 20:07 < kanzure> hrm, anyway 20:07 < fenn> what you trying to do overall? 20:08 < kanzure> we're making the repressilator component in the diagram on pg 7 20:08 < kanzure> that's in the lab 20:08 < kanzure> now, what I want to do re: your "shine a light to synthesize DNA" system 20:08 < kanzure> the lab project allows that 20:08 < fenn> oh, i thought you were just pulling a repressilator off the shelf and using it for someting else 20:08 < kanzure> nah 20:09 < kanzure> as I was saying - the light frequency input - can be done via this method. So if you can program the state machine with pulsing photons, then the synthesizing component needs to come into play to do it nucleotide-by-nucleotide with the following requirements: 20:10 < kanzure> (1) not interfering with the existing template/non-template switch and source strands 20:10 < kanzure> (2) lots of nucleotides flying around ready to be used to make the next piece of the DNA strand 20:10 < kanzure> (3) looks like 2 bits of memory? since we need 4 different nucleotide possibilities? 20:10 < kanzure> but then we need some more data, actually 20:10 < kanzure> since there's also stuff like how to end the molecule 20:10 < kanzure> (it'll just be controlling the oligonucleotide reaction process) 20:11 < kanzure> really, I'm hoping that oligo synthesis could work with these transcriptional elements nearby, otherwise the idea's foobarred 20:11 < fenn> wtf are you talking about memory? 20:11 < fenn> and state machines? 20:12 < kanzure> did you look at the diagram? 20:12 < fenn> no, where is that? 20:12 < kanzure> email, in the pdf page 7 20:12 < kanzure> top diagram 20:13 < kanzure> I'm not sure if we actually need 'memory' for this 20:13 < kanzure> if we need more than 1 bit, then yes 20:13 < kanzure> since to do an adenosine versus guanine you'd have to do 2 pulses, etc. 20:14 < kanzure> 4 nucleotides, I mean 20:14 < kanzure> so that's at least 2 bits 20:14 < kanzure> and then there's a few other details that might have to be implemented ... like the way to end the strand or something? I'm not entirely versed in oligo synthesis parameters. 20:14 < kanzure> or the variations to the protocol, I mean. 20:15 < fenn> yep 4 nucleotides needs 2 bits, but i'd settle for 2 nucleotides 20:16 < kanzure> how so? 20:16 < fenn> uh, nevermind 20:16 < kanzure> in terms of awesomeness, or a way to 20:16 < kanzure> yeah 20:16 < fenn> i dont think there exist any enzymes to synthesize a dna or rna strand from nothing at all 20:17 < fenn> i'm thinking something involving tRNA.. not sure why 20:19 < kanzure> wait, this is going to suck 20:19 < kanzure> transcriptional switches take forever 20:19 < fenn> yes 20:19 < kanzure> but it might be faster than building a DNA synthesizer 20:19 < kanzure> would it? 20:20 < fenn> if you have a dna lab to work in :0 20:20 < kanzure> oh 20:20 < kanzure> re: 2 nucleotides 20:20 < fenn> uh, if you get it working it would be worth zillions 20:20 < kanzure> just have two different colonies 20:20 < kanzure> each with 2 bit input 20:20 < kanzure> i.e., light/dark maybe 20:20 < kanzure> and then integrate or something 20:20 < kanzure> hrm 20:20 < kanzure> that wouldn't work 20:21 < kanzure> (zipper strategy) 20:21 < fenn> i was thinking 3 wavelengths of light 20:21 < kanzure> one colony does A & T, the other does G & C. but the zipper issue ... lineup, etc ... 20:21 < kanzure> well, light/dark is easy because you can just cover it up or not 20:21 < fenn> but that doesnt fit the ellingtonia agenda 20:21 < fenn> need to squeeze transcriptional switches into the design somewhere :) 20:21 < kanzure> heh 20:21 < fenn> could build a demultiplexer out of trans switches 20:22 < kanzure> basically I'm ok with needing multiple light sources, but it needs to be something doable on the cheap 20:22 < kanzure> for example, building fires that burn different colors 20:22 < kanzure> like distinct colors. 20:22 < fenn> uhh whats wrong with an LED? 20:22 < kanzure> can you make an LED? 20:22 < fenn> no 20:22 < kanzure> can you bash rocks together to make sparks? 20:22 < fenn> ok i see where you're going 20:22 < kanzure> :) 20:23 < fenn> hmm. what kind of tech level is allowed? 20:23 < kanzure> that's a good question 20:23 < kanzure> none would be awesome 20:23 < kanzure> but I'm willing to compromise 20:23 < fenn> fluorescent tubes give off different kinds of light 20:23 < fenn> depending on what gas is in it 20:23 < kanzure> the way that I seem to be working here is starting with 'advanced tech' and then finding a simplification of it 20:23 < kanzure> vacuum tubes? 20:23 < fenn> geissler tubes 20:24 < kanzure> we talked about these before methinks 20:24 < kanzure> glassblowing sucks 20:24 < kanzure> or blows 20:24 < fenn> just a glass tube filled with gas and HV 20:24 < fenn> anyway it's 1800's tech 20:24 < kanzure> but if necessary, glassblowing is acceptable 20:24 < fenn> i'm trying to think of a way to convert electrical signals to light signals using biological materials 20:24 < kanzure> oh, so different gases burn different colors, right? 20:24 < kanzure> uhh 20:25 < fenn> like frankenstein firefly 20:25 < fenn> bzzt bztzt 20:25 < kanzure> so, organic LEDs were in the news recently 20:25 < kanzure> fluoroproteins 20:25 < kanzure> now we need a bioluminescent, electrical protein 20:25 < kanzure> what do fireflies use anyway? 20:25 < fenn> luciferase 20:25 < kanzure> enzyme reaction? 20:26 < kanzure> 'Light production in fireflies is due to a chemical reaction that occurs in specialized light-emitting organs, usually on the lower abdomen. The enzyme luciferase acts on luciferin in this organ to stimulate light emission.' 20:26 < kanzure> surely different fireflies do different colors? 20:26 < fenn> its activated by oxygen, the blinking is due to their flying up/down cycle which increases respiration 20:26 < kanzure> hell, let's do a selection experiment on firefly light color 20:26 < kanzure> wouldn't be too hard to set up a firefly cage 20:26 < fenn> i've only seen green fireflies 20:26 < kanzure> surely it's mutatable? 20:26 < kanzure> aha 20:27 < kanzure> ellingtonia specializes in knocking out nucleotides in genes 20:27 < fenn> i know there are blue and red light-emitting proteins in deep sea creatures 20:27 < kanzure> with unnatural nucleotides 20:27 < fenn> even infrared 20:27 < kanzure> which significantly significantly enhances mutation 20:27 < kanzure> mutation rates, I mean 20:27 < kanzure> not mutation results 20:27 < kanzure> oh, there's also the jellyfish stuff 20:27 < kanzure> GFB, luciferase/luciferin, ? 20:27 < kanzure> would really like a solid state laser ... but luciferin would be totally solid state 20:28 < kanzure> blah, lots of labs have a list of fluoroproteins right? 20:28 < kanzure> not focused/directed -- how much energy loss, noise, etc.? 20:28 < kanzure> holding up a firefly is not like holding up a lamp 20:28 < kanzure> maybe we can try making a huge mass of firefly luciferin 20:28 < kanzure> as big as we can make it 20:29 < kanzure> and just dump a huge bucket of luciferase on it and see if it lights up like a firecracker :) 20:29 < fenn> sure, or just a couple drops 20:29 < kanzure> how are we going to get that much luciferin 20:29 < fenn> standard recombinant dna stuff 20:29 < kanzure> it would require its own experiment before we even confirm this method though 20:30 < fenn> to see if combining luciferin/luciferase generates light (quickly)? 20:30 < kanzure> if it's accumulative 20:30 < fenn> accumulative? 20:30 < kanzure> i.e., will 5 grams of luciferin light up strong enough to signal the transcriptional circuit? 20:30 < fenn> ah 20:30 < fenn> probably not 20:30 < fenn> :( 20:31 < kanzure> bio is the only alternative to solid state that I'm going to accept here, I think. not sure. 20:31 < kanzure> solid state advantages: it's freaking reliable (photons, lasers) 20:31 < kanzure> bio advantage: just replicate more 20:31 < fenn> solid state is also quite cheap these days 20:31 < kanzure> replicate enough to do the synthesis you're planning for a few weeks in advance, or something 20:31 < kanzure> also, DIY laser projects on the net 20:31 < kanzure> and organic LEDs as another option.. 20:32 < kanzure> but on the other hand, "go build a fire with sticks and stones" might work, right? 20:32 < kanzure> http://en.wikipedia.org/wiki/Organic_light-emitting_diode 20:32 < kanzure> oLEDs might be too manufacturing-specific (i.e., fabs-only) 20:33 < fenn> so are LED's 20:33 < kanzure> wouldn't it be nice if we could just have a pigment filter for a light source 20:33 < kanzure> then we just need to figure out how to get enough candles / luminosity 20:33 < kanzure> or some laser variable. 20:33 < kanzure> intensity 20:33 < fenn> what about sunlight/colored gel/shutter 20:33 < kanzure> colored gel? 20:34 < kanzure> hrm 20:34 < fenn> to filter out the light 20:34 < kanzure> oh, food coloring 20:34 < kanzure> well shit. 20:34 < fenn> its called a gel for some reason 20:34 < kanzure> food coloring works very, very well 20:34 < kanzure> it blocks out specifically that color 20:34 < kanzure> so it's the absence of it 20:34 < kanzure> that's ... close. 20:34 < kanzure> would it be enough? 20:34 < fenn> the thing is, your eye sees only one color at a time 20:34 < fenn> but we really want monochromatic light 20:34 < kanzure> hm? 20:35 < fenn> and a messy spectrum with everything but green missing will look red 20:35 < fenn> even though its not monochromatic red 20:35 < kanzure> how is that possible? 20:35 < fenn> er.. with green missing i mean 20:35 < kanzure> everything-but-green ==> red? 20:35 < kanzure> oh 20:35 < kanzure> green-missing => red? 20:35 < fenn> yeah 20:35 < kanzure> is this something that I should be aware of already 20:36 < kanzure> like 7th grade science or something? sounds very peculiar 20:36 < fenn> hmm.. you know rods/cones, yah 7th grade science 20:36 < fenn> look up color blindness 20:36 < kanzure> okay 20:36 < kanzure> are we talking about the human eye, or color input for any system 20:37 < fenn> i'm just saying food coloring won't necessarily produce a tight spectrum 20:37 < kanzure> oh, pigmentation purity issues 20:37 < kanzure> is that it? 20:37 < fenn> how about a prism or diffraction grating 20:37 < kanzure> http://en.wikipedia.org/wiki/Diffraction_grating 'In optics, a diffraction grating is an optical component with a regular pattern, which splits light into several beams travelling in different directions. ' (huh, like a prism) 20:37 < kanzure> hrm. 20:38 < kanzure> crystals? 20:38 < kanzure> how does one go about making a prism 20:38 < kanzure> http://en.wikipedia.org/wiki/Prism_(optics) 20:38 < fenn> a triangular chunk of something clear 20:38 < kanzure> even gel? 20:39 < fenn> sure 20:39 < kanzure> can we grow gels? 20:39 < fenn> i dont know how you'd get it into a triangular shape 20:40 < kanzure> if it's stable, then physical sawing / sharp pointy stick 20:40 < kanzure> how exact? 20:40 < fenn> who's gonna saw it? 20:40 < kanzure> me? 20:40 < fenn> oh 20:40 < fenn> in that case, why make it from gel 20:40 < kanzure> oh, what if the gel grows in a triangular container? 20:41 < fenn> you still need a triangular container 20:41 < kanzure> actually, I've heard some terms recently 20:41 < kanzure> apparently you prepare a gel in the lab 20:41 < kanzure> by mixing some chemicals together 20:41 < kanzure> are these chemicals, biological in origin? 20:41 < kanzure> an ECM of some sort? 20:41 < fenn> ECM? 20:41 < kanzure> extracellular matrix 20:41 < kanzure> re: needing the container; you need the tank too. 20:42 < fenn> ok, where do you get the tank? 20:42 < kanzure> scrap metal :( that's an unsolved problem 20:42 < fenn> anyway, there's a big ugly complex electromechanical shutter mechanism to worry about too 20:42 < fenn> granted, it could be just a motor with some fan blades 20:43 < kanzure> easy system: cardboard paper + servo motor to spin it around + keep it in the path from light -> circuit; then, spin it around to open the gate 20:43 < fenn> but those dont exactly grow on trees either 20:43 < kanzure> yes 20:43 < kanzure> right 20:43 < kanzure> well, we mentioned an electrical shock 20:43 < kanzure> which was to be used to do the bioluminescence 20:44 < kanzure> in this case, however, we just need a shock to clear a 'buffer' or 'blocker' 20:44 < kanzure> hrm, that sucks 20:44 < fenn> you're thinking like a liquid crystal? 20:44 < fenn> or a kerr cell (ignore) 20:44 < kanzure> liquid crystal might work, how would we implement? 20:45 < fenn> i think this project is somewhat lacking in rigorous specifications 20:45 < kanzure> of course 20:47 < fenn> i think its best to ignore the 99% of the rest of the technological infrastructure leading up to getting blinky lights 20:47 < fenn> and just pretend you've got blinky lights 20:48 < kanzure> oh 20:48 < kanzure> I guess somebody else can come along and do it anyway 20:48 < kanzure> the blinky lights aren't too important to the overall circuit really 20:48 < fenn> the thing is, we dont have a direct write method for DNA yet 20:48 < kanzure> remember, the frequency-input mechanism on the bio side isn't necessarily light-based 20:49 < kanzure> it might be electrical if we want, or something, I'll have to ask the guys at the lab about this 20:49 < kanzure> okay, so the DNA write method 20:49 < fenn> ok 20:49 < kanzure> http://heybryan.org/mediawiki/index.php/DNA_synthesizer 20:49 < kanzure> I have some notes on the oligo synthesis reactions, but not much 20:49 < kanzure> starting @ "The first step in the synthesis is the removal of the dimethoxytrityl (DMT) group with TCA to free the 5'-hydroxyl for the coupling reaction. The next step, activation, is achieved by simultaneously adding a phosphoramidite derivative of the next nucleotide and tetrazole, a weak acid to the reaction chamber." 20:50 < kanzure> "The tetrazole protonates the nitrogen of the phosphoramidite, making it susceptible to nucliophilic attack. This intermediate is very reactive and the following coupling step is complete in less than 30 seconds. The 5'-OH group of the phosphoramidite is blocked with the DMT group." 20:50 < fenn> blah 20:50 < fenn> i dont care 20:50 < fenn> that's chemical engineering bullshit 20:51 < kanzure> but how would we build a DNA molecule 20:51 < kanzure> if we could make it think that it's building the complementary strand, that'd be great 20:51 < kanzure> it == transcription factors and so on 20:51 < kanzure> erm, proteins related to the proc 20:51 < fenn> you'd have one molecule to hold the base in place, and then ligase comes along and bonds it to the backbone 20:52 < fenn> know how a line printer works? 20:52 < kanzure> ink based? 20:52 < kanzure> I know how a laser jet works 20:53 < kanzure> the 'laser' fires and writes to the roller, the roller presses down on the paper 20:53 < fenn> there's a wheel which spins around to the correct letter (nucleotide) and then it goes *wham* and hits against a piece of paper (ligation) 20:53 < fenn> you can have a whole bunch of wheels spinning in parallel, so you print a whole line at a time 20:54 < fenn> anyway, i'm thinking just one wheel 20:54 < fenn> with 4 letters 20:54 < kanzure> ok, when you say wheel that spins around to the correct nucleotide, I'm thinking casino "777 YOU WIN!!" spinners 20:54 < kanzure> and it sounds like a nasty biomolecule ... 20:54 < fenn> yeah something like that 20:55 < fenn> actually it doesnt even need to ligate, just hold them in place while another molecule uses them as a template 20:55 < fenn> dunno if that would really work though 20:55 < kanzure> I don't like this. sounds like spooky molecular engineering. 20:55 < kanzure> and the molecule *must* be replicable (i.e., be a protein) 20:55 < fenn> ya 20:56 < kanzure> hey, uh, to speed up transcriptonial switches, why not run them while in a centrifuge? 20:56 < fenn> nope 20:56 < fenn> speed is governed by concentration and temperature 20:56 < kanzure> it's just diffusion that limits it, right? 20:56 < kanzure> concentration doesn't matter if they're covering the same space more quickly 20:57 < fenn> centrifuge just increases the pressure, which causes some equilibrium shifts (favoring larger molecules) 20:57 < fenn> also it wreaks havoc with molecular machinery because everything gets clumped together at the bottom 20:57 < kanzure> sucks 20:58 < kanzure> I don't like your biomolecule idea really 20:58 < kanzure> what we have is some molecular gates that can hold the 'state' of the system 20:58 < kanzure> I mean, hold which nucleotide needs to be done next 20:58 < fenn> so, in computer terms that's a register 20:58 < kanzure> (there is obvious noise issues -- some might still be manufacturing the last state's nt, but just encode some selection experiment into it) 20:58 < kanzure> yes 20:58 < kanzure> so if you have a register, what do you connect to 20:59 < kanzure> to specify the next nucleotide? 20:59 < kanzure> usually the only way that the next nucleotide is specified is ... mutation 20:59 < kanzure> I know it sounds weird, but think it through 20:59 < kanzure> you have a dna molecule and the only way you change it is through mutation 20:59 < kanzure> or recombination with another dna molecule 20:59 < kanzure> (so we need a ghost dna molecule) 20:59 < kanzure> anyway, so the 'next' nucleotide is basically set unless the machinery itself makes an error 21:00 < kanzure> so we need to be able to engineer these errors to be subject to our will 21:00 < kanzure> via the signaling system we have 21:00 < kanzure> make sense? 21:00 < fenn> hmm.. could use a genengineered DNA repair enzyme to 'repair' each nucleotide incorrectly (swap in what we want) 21:00 < kanzure> how would it know what spot it is at 21:00 < kanzure> also, dna repair enzymes do not move so slowly 21:01 < fenn> perhaps it could be light activated to move forward one bp 21:01 < fenn> and otherwise do nothing 21:01 < fenn> so "prep - load guanine" then "swap" 21:01 < kanzure> movement kinetics ? 21:01 < kanzure> hrm 21:01 < kanzure> needs to have a piece of the 'protein puzzle' missing before it can move 21:02 < fenn> not so much different from what i was saying before, only now it's crawling along a dna strand 21:02 < kanzure> and then we swap in the replacement component via light signaling (or whatever) 21:02 < kanzure> but the problem is that the concentrations suck 21:02 < fenn> oh, you can use a single stranded dna and just fill in the blanks 21:02 < kanzure> hm? 21:02 < kanzure> no no 21:02 < kanzure> I mean the repair enzmye needs to have a "push here to go" button 21:02 < kanzure> and we provide the push via a particle 21:03 < kanzure> however, these particles are everywhere 21:03 < fenn> i dont understand 21:03 < kanzure> who's to say that another particle won't hit that same enzyme a few seconds later? 21:03 < kanzure> and therefore make it do two bases when we only really meant one? 21:03 < fenn> that's what i'm trying to avoid, hence the "swap" signal 21:03 < kanzure> the problem is that the DNA repair enzyme only works at a certain rate, and we need it to be 'phase locked' to our switch signals 21:03 < kanzure> hm 21:03 < kanzure> how would the swap signal avoid that? 21:03 < kanzure> oh, the signal is extra 21:03 < kanzure> another extra signal on top of it all? 21:04 < fenn> you'd give it enough time to get ready (bind to the correct nucleotide) before giving the swap signal 21:04 < kanzure> (would be nice if the DNA repair enzyme had a transcriptional switch attached to it .....) 21:04 < fenn> another signal (wavelength of light) 21:04 < kanzure> okay, so the transcriptional circuit would be used to inhibit/activate the production of the first signal 21:04 < kanzure> and then another light type for the swap signal after a certain amount of time 21:04 < kanzure> (that amount of time is to let the switches do their thing and so on) 21:05 < kanzure> the 'first signal' would specify what the molecules are looking for 21:05 < fenn> i'm not so fond of transcriptional circuits really 21:05 < fenn> i guess you could use the oscillator as a clock/swap signal 21:05 < kanzure> I don't think you understand though .. it's all DNA 21:05 < fenn> but it would have to be predictable to within some very small percent 21:05 < fenn> or you'll get framing errors 21:05 < kanzure> well, that's a flip-flop of what I was thinking of. I wanted the swap signal to be some light source; and then the other one to be due to switches 21:06 < fenn> if the other one is due to switches, you have no data 21:06 < fenn> just some silly on-off pattern 21:06 < fenn> 01010101010101 21:06 < kanzure> remember the 2 bits? 21:06 < kanzure> oh 21:06 < kanzure> but if you have 2 bits then you could select different particles to send the enzymes 21:06 < kanzure> the repair enzymes that are waiting for the 'signaler' as for what to fetch ext 21:06 < kanzure> *next 21:07 < kanzure> wait, is it a "what to fetch next signal" ? so the two signals are officially: (1) fetch-next, (2) swap 21:07 < kanzure> *next fetch 21:07 < kanzure> fetch and swap 21:07 < fenn> if i weren't so lazy i'd make a blender animation of this thing 21:07 < kanzure> if I didn't suck so much with blender, I'd make an animation of this thing 21:08 < fenn> yeah, that too 21:08 < kanzure> okay, so the fetch signal -- it would be really, really great if this was an switch template from the synthetic circuits 21:08 < kanzure> Ellington has supposedly worked with mutated polymerases 21:08 < kanzure> or invented them or something :) 21:08 < kanzure> is it polymerase that we want? 21:09 < fenn> well.. polymerase can have dozens of pieces, each with their own function 21:09 < fenn> some polymerases are simpler than others 21:09 < fenn> dna repair polymerase, for example pol-I, are usually simpler 21:10 < kanzure> I don't like the 'binding complex on polymerase' thing ... seems kind of hard to engineer, since polymerase is, itself, a protein 21:10 < kanzure> (for the fetch signaler) 21:10 < kanzure> erm, so we need multiple pol-I molecules floating around 21:10 < fenn> check 21:10 < kanzure> different ones for different molecules, we don't want it to have to have internal logic 21:10 < kanzure> not a big issue, just a note 21:10 < fenn> hmm 21:11 < kanzure> without the fetch signaler it shouldn't work 21:11 < kanzure> but when it has the fetch signal, only those that fit the fetch signal will work 21:11 < fenn> pol-I crawls along the dna strand one base at a time 21:11 < kanzure> crap 21:11 < fenn> that's useful because it means you dont lose your place 21:11 < fenn> also, its guaranteed that the enzyme will be there 21:11 < kanzure> ok, scratch that last stuff about multiple types of pol-I then 21:11 < fenn> so you only have to wait until a free nucleotide comes along 21:12 < kanzure> we can't flush it quickly enough 21:12 < kanzure> what are the specs here? 21:12 < fenn> reliability 21:12 < kanzure> needs to be able to accept a fetch and swap signal 21:12 < kanzure> is that about it? 21:12 < fenn> yep 21:12 < kanzure> the fetch signal will be any of 4 types 21:12 < kanzure> the swap signal will tell it to move 21:13 < kanzure> hm, doesn't it move thanks to ATP or something? 21:13 < kanzure> I don't recall. 21:13 < fenn> the nucleotide has "ATP" built into it 21:13 < kanzure> http://en.wikipedia.org/wiki/DNA_polymerase_I 21:13 < kanzure> oh 21:13 < fenn> ATP is just adenosine with phosphate sticking off it 21:13 < fenn> there's also GTP TTP CTP 21:14 < kanzure> http://ellingtonlab.org/ ' For example, we have evolved RNA polymerases that can utilize modified nucleotides. Finally, we have extended our evolutionary approaches to whole organisms, and are attempting to evolve 'unnatural' E. coli (unColi) that can augment their genetic codes with unnatural amino acids.' 21:14 < kanzure> so evolving RNA polymerase seems to be possible 21:14 < kanzure> grr 21:14 < kanzure> how do we evolve polymerase for such specific functions? 21:15 < kanzure> (fetch, swap) 21:15 < fenn> i dont see the point of that really 21:15 < kanzure> increases rate of mutation 21:15 < kanzure> thanks to 'new' nucleotide types 21:15 < fenn> how do they specify the location of the new nucleotide? 21:15 < kanzure> thus you can put some intense selective pressure on them 21:15 < kanzure> hrm 21:15 < kanzure> maybe he's not? 21:15 < fenn> there's lots of ways to increase mutation 21:15 < kanzure> radiation hurray 21:16 < fenn> location-specific mutation is useful. also highly random mutation. most methods are somewhere in between 21:16 < kanzure> ''DNA replication in E. coli proceeds at approximately 1,000 nucleotides/second, 21:16 < fenn> anyway, it doesnt get us any closer to in-vivo DNA synthesis 21:17 < kanzure> ' while the rate of synthesis by pol I averages only 20 nucleotides/second. ' 21:17 < kanzure> v 21:17 < kanzure> http://en.wikipedia.org/wiki/DNA_polymerase_I 21:17 < fenn> DNA replication is highly optimized for speed, accuracy, and efficiency. not necessarily simplicity or hackability 21:17 < fenn> 20bp/s is good enough 21:17 < fenn> e. coli genome is like 4kbp right? 21:18 < kanzure> wtf? sounds unlikely 21:18 < kanzure> but useful 21:18 < fenn> one would expect this to be an easy google search 21:19 < kanzure> 5E6 21:19 < kanzure> http://www.genome.wisc.edu/aboutus.htm 21:19 < fenn> hmm ok Mbp 21:19 < kanzure> 4000 genes 21:19 < kanzure> so you were off by a term 21:19 < fenn> 12 hours 21:20 < fenn> oop, 70 hours 21:20 < kanzure> ok, at 20 nt/sec, that has to be incredibly in sync 21:20 < fenn> wah 21:20 < kanzure> unless we can slow it down 21:20 < kanzure> hey, let's slow it down with a signal too 21:20 < kanzure> let's throw in another light source 21:20 < fenn> why is it slow in the first place? 21:20 < fenn> ok, premature optimization, who cares how fast it is 21:21 < kanzure> Paula de Lucia; John Cairns (Dec 1969). "Isolation of an E. coli Strain with a Mutation affecting DNA Polymerase". Nature 224 (5225): 1164-66. doi:10.1038/2241164a0. PMID 4902142. 21:21 < kanzure> well, we do 21:21 < fenn> december 1969, when everything happened 21:21 < kanzure> because it has to be slow enough to let the switches do their magic :( 21:21 < kanzure> although 21:21 < kanzure> hrm 21:22 < kanzure> the switches obviously complicate things, but I'm sure there's a way 21:22 < kanzure> generally halting it is a good thing 21:22 < kanzure> has to be light, there's no way a diffuse reactant can get to all of them in time 21:23 < fenn> could induce strand separation by coding a bunch of C-G's 21:23 < kanzure> huh? 21:23 < fenn> to stop the enzyme, kick it away from the template strand 21:23 < kanzure> so pad everything? 21:23 < fenn> C-G binding is weaker than AT binding 21:23 < kanzure> how would the sequence be 'recovered'? 21:23 < kanzure> oh 21:23 < fenn> see what i'm saying? 21:23 < kanzure> sort of 21:24 < kanzure> no, I don't see how it would work to our advantage 21:24 < fenn> ok nevermind 21:24 < kanzure> ATGCCTAGGTCGCGCGCGC 21:24 < kanzure> but now you have C-G's 21:24 < kanzure> instead of your sequence :( 21:24 < fenn> right 21:25 < kanzure> and if you endonuke it, too many fragments right? 21:25 < fenn> you can predict where restriction enzyme will cut 21:25 < kanzure> oh, so the buffer will be the sequence for endonuclease 21:25 < kanzure> then you do an endonuclease cut 21:26 < fenn> yeah they are very specific to certain sequences (5 bp or so) 21:26 < kanzure> 21:26 < kanzure> then endonuke it 21:26 < kanzure> but you lose the position 21:26 < kanzure> wasn't that the point of polymerase? 21:26 < fenn> isnt that the point? 21:26 < fenn> you're done 21:26 < kanzure> no 21:27 < kanzure> that's for one base 21:27 < kanzure> :( 21:27 < fenn> oh, why kick the polymerase off? 21:27 < kanzure> to stop the enzyme 21:27 < fenn> wait, are you saying use transcriptional circuits to synthesize a new polymerase enzyme for each base? 21:27 < kanzure> so that you can give the switches time to induce the building of the 'fetch' signaler molecules 21:27 < kanzure> nono 21:28 < kanzure> the circuits control the production of the 'fetch-next' molecules 21:28 < fenn> ok so the 'fetch' molecules are something like tRNA 21:28 < kanzure> shit 21:28 < kanzure> that sounds right :) 21:28 < kanzure> \o/ translation? 21:28 < fenn> well, tRNA normally grabs amino acids 21:28 < fenn> yes 21:29 < kanzure> tRNA could be an easy byproduct of the switches, I think 21:29 < fenn> or it could be a short aptamer 21:29 < kanzure> how would you dehook whatever the aptamer catches, (dehooking -> give to polymerase) 21:30 < fenn> i dont see the need for a separate fetch molecule? 21:30 < kanzure> funny, Wikipedia page "Translation" is on linguistic translation. Even though bio translation occurs in trillions of cells *per organism*. 21:30 < kanzure> well, the fetch molecule tells the polymerase what to fetch next 21:30 < fenn> so its more like information transfer than molecule transfer 21:30 < kanzure> sure 21:30 < kanzure> the actual molecule doesn't matter 21:31 * fenn pictures floppy disks being shuffled around 21:31 < kanzure> (but the swap signal, that probably should be a molecule or something) 21:31 < kanzure> huh? 21:31 < kanzure> that's nearly before my time 21:31 < fenn> hah 21:31 < kanzure> but I do remember loading floppies into machines 21:31 < kanzure> to install software. 21:31 < kanzure> "insert next disk" 21:32 < fenn> 'never underestimate the bandwidth of a station wagon loaded with tapes hurtling across the countryside' 21:32 < kanzure> unless you refer to the concept of a stack/list/queue 21:32 < kanzure> so we need to get the information to the polymerase molecule 21:33 < kanzure> and we need it to act on this information somehow 21:33 < kanzure> kind of like switching out a prosthetic hand 21:33 < kanzure> a new hand for a different job (eating, versus playing catch with a kid) 21:33 < kanzure> (I've heard that people used to do that with old prosthetic hands) 21:34 < fenn> prosthetic hands is a pretty sad state of affairs 21:34 < fenn> so, instead of switching hands and then looking around for them each time, why not have a "wheel" with all the hands you need on it already 21:35 < kanzure> how do you tell the wheel to rotate, and how do you tell it to rotate a certain amount, and how do you know that you're not rotating the wheel multiple times? 21:35 < fenn> the wheel rotates on its own, but gets locked into a certain position when some "solenoid" is on 21:35 < fenn> except in this case the solenoid is a light-activated protein 21:36 < fenn> so when there's no input, the wheel is just spinning due to brownian motion 21:37 < kanzure> how do you make a molecular wheel 21:37 < fenn> or, it could be any old binding protein that gets synthesized by your transcriptional switch 21:38 < fenn> i suppose it doesnt need to be a wheel, just some free floating proteins in close proximity, held together with 'chains' 21:38 < fenn> so you just write it all on one gene like we talked about earlier with the purification tag 21:39 < fenn> the fetch molecule would then come along and pull everything together 21:39 < kanzure> wha? 21:39 < kanzure> chains? 21:39 < fenn> ok, so, instead of throwing your prosthetic hand behind the couch after you're done eating, you keep all your hands tied to your belt 21:40 < kanzure> do we have a belt? 21:40 < fenn> yeah, its the polymerase 21:40 < kanzure> it doesn't seem to work like that 21:40 < fenn> "you" are the polymerase 21:40 < kanzure> I mean, it has no reconfigurable 'hand joint' 21:40 < fenn> wah 21:40 < kanzure> i.e., in the prosthetic hand case, "you are the polymerase" -> "you're the guy with the hook for the hand" 21:40 < kanzure> how would you even think about evolving that ? 21:40 < kanzure> the reconfigurable joint 21:41 < fenn> i'm gonna go look for some youtube animations of polymerase 21:41 < kanzure> how does it do it anyway? 21:41 < kanzure> yeah 21:42 < fenn> i'm hoping for a nice animation like for atp synthase, but its unlikely 21:43 < kanzure> needs a 'hand eject' function 21:45 < fenn> this is cool, a little too fast though :) http://youtube.com/watch?v=49fmm2WoWBs&feature=related 21:47 < fenn> also very nice http://www.youtube.com/watch?v=E8NHcQesYl8&NR=1 21:49 < kanzure> need to have a 'hand eject' function 21:49 < kanzure> think of it as if you're causing neurotransmitter receptor sites in plasma membranes to stop working 21:49 < kanzure> erm, to just float off 21:49 < kanzure> like lifeboats 21:50 < kanzure> problem with that is like asking for you to detach an ion channel :) an ion channel is a *channel* :) not a flagella-like molecular extension 21:56 < fenn> another cool animation (ribosome) http://www.youtube.com/watch?v=Jml8CFBWcDs&feature=related 21:58 < kanzure> http://scholar.google.com/ search 'polA polymerase reduced OR slow OR inhibited' 21:58 < kanzure> 'DNA polymerase active site is highly mutable: Evolutionary consequences' 21:59 < kanzure> http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=25787 21:59 -!- ybit [n=u1@unaffiliated/ybit] has quit [Read error: 104 (Connection reset by peer)] 21:59 -!- h2i [n=u1@unaffiliated/ybit] has joined #hplusroadmap 22:00 < kanzure> ' (motif A) of Thermus aquaticus DNA polymerase I. After selection, by using genetic complementation, we obtained a library of approximately 8,000 active mutant DNA polymerases, of which 350 were sequenced and analyzed. This is the largest collection of physiologically active polymerase mutants. We find that all' 22:00 < kanzure> woah :) 22:00 < fenn> meh 22:00 -!- h2i [n=u1@unaffiliated/ybit] has quit [Client Quit] 22:00 < fenn> patent avoidance search :) 22:01 < kanzure> ' Bacteria dependent on these mutated polymerases for survival are fit to replicate repetitively. ' 22:02 < kanzure> hellooo 22:02 < kanzure> ' that bears the temperature-sensitive DNA polymerase I mutatation, polA12. The mutant enzyme has a reduced electrophoretic mobility and sedimentation rate. It is abnormally thermolabile and is rapidly inactivated at low salt concentrations.' 22:02 < kanzure> salt is metallic, use quick magnets 22:04 -!- Vedestin [n=Vedestin@d122-109-35-58.sbr3.nsw.optusnet.com.au] has joined #hplusroadmap 22:06 < fenn> what we need is 80/20 for proteins 22:06 < kanzure> ? 22:06 < fenn> some basic folding motif that we can plug together like legos to build molecular structure 22:07 < kanzure> folding chemistry consists of what? 22:07 < fenn> instead of custom holistically-crafted kanji characters 22:07 < fenn> something more like the roman alphabet 22:07 < Vedestin> changing the physical structure of proteins 22:08 < kanzure> Vedestin: would require protein folding, or lots of selection 22:08 < kanzure> fenn is talking about something, else 22:08 < Vedestin> oh is he? 22:08 < kanzure> yes, 'instead of custom holistically-crafted kanji characters' 22:08 < fenn> i'm trying to think of an analogy 22:08 < kanzure> I thought the language analogy was mostly good 22:09 < kanzure> but kanji is still object/pattern recognition 22:09 < kanzure> there's nothing quite like the way that life is doing it 22:09 < kanzure> well 22:09 < kanzure> there has to be, it's physically occuring 22:10 < fenn> the idea is to build a discrete set of modules that stick together, and design your protein machine based on those instead of trying to fold the whole thing like origami 22:11 < kanzure> probably avoiding protein folding 22:11 < kanzure> or if we do folding, it's only through mechanisms we know about 22:12 < fenn> it throws the pol-I out with the bath water though 22:12 < kanzure> of course 22:12 < kanzure> it throws mostly everything out 22:12 < fenn> i.e. i still dont know how to make a 'hand' 22:12 < fenn> so, i think this is more than an evening's work 22:13 < kanzure> well, we just need to hack polA 22:13 < kanzure> pol I, whatever 22:13 < kanzure> so I'm thinking a few ways: (1) selection experiments, mutations 22:14 < kanzure> (2) biomolecular engineering with protein folding. let's hope we can model the folding of polymerase. 22:14 < kanzure> and then we can do something computational like that 22:14 < fenn> hack it how? 22:14 < kanzure> we're obviously going to make a new polA, right? 22:14 < kanzure> so we'll take the sequence and modify it 22:15 < fenn> um, yes, but how do we know the correct sequence 22:15 < kanzure> '… response of E. coli to low levels of alkylating agent: The role of polA in killing adaptation ' 22:15 < kanzure> right, we have to figure out how to 'hack' it 22:15 < kanzure> I bet somebody has mapped this 22:15 < fenn> it's the old 'protein folding' problem 22:15 < kanzure> let's say it has a seq of len 5 22:15 < kanzure> of len 50, actually 22:15 < kanzure> so maybe bp 5 to 15 are what we are looking for 22:15 < kanzure> if so, we're in super-duper luck 22:16 < kanzure> the smaller the better :) 22:16 < kanzure> we're looking for some way to control mutagenesis, remember 22:16 < fenn> no, i could care less about mutagenesis 22:16 < kanzure> so we need to be able to leave some sort of 'molecular hookon' jiggler that we can trigger when we have the polymerase molecule going about 22:16 < kanzure> erm? the idea is to control when it 'makes an error' 22:17 < kanzure> to hack the error process 22:17 < kanzure> and just make that the default 22:17 < fenn> it should always be making an error 22:17 < kanzure> and be in control (grab life by the horns) 22:17 < kanzure> heh, paper title: The Horns of Polymerase 22:17 < fenn> ok, we have good control over the sequence, but no way to map that to a folded structure except through reverse engineering 22:18 < kanzure> right, but lots of people have focused on polymerase already 22:18 < kanzure> so we might have a 'map of polymerase mutations' 22:18 < kanzure> and if anything on that map looks potentially useful 22:18 < kanzure> then we're good 22:18 < fenn> if we had a direct write tool, we could do lots of trial and error to figure out some motifs 22:18 < kanzure> right 22:18 < kanzure> so it might be one-to-one like that in the case of polymerase, it might not be 22:18 < kanzure> it's something to investigate in the literature 22:18 < fenn> but we dont have a direct write tool, so its kind of a dependency loop 22:18 < kanzure> I have a lab 22:19 < fenn> i think modern methods are too slow and labor intensive to get anywhere in a reasonable amount of time/money 22:19 < fenn> $15/bp? well what if it takes a couple billion bp in trial/error to figure out some basic patterns 22:20 < fenn> so, how many bp does it take anyway? :) 22:20 < kanzure> $0.15/bp http://e-oligos.com/ 22:20 < kanzure> hrm 22:20 < kanzure> we would have to evolve it without sequence engineering :( 22:21 < kanzure> and I'm clueless as to the design of selection experiments, where we want to add input/output 22:21 < kanzure> to the target.. 22:21 < fenn> veolve what? 22:21 * fenn kicks ssh in the nads 22:21 < kanzure> evolve our hacked polymerase 22:21 < kanzure> oh, lag? 22:22 < fenn> ping works without any delay, but for some reason my typing doesnt get echoed for like 20 seconds 22:22 < kanzure> uh, also, how would we make polymerase work anyway? it needs a template strand to work with 22:23 < kanzure> a template strand to copy, I mean. 22:23 < fenn> u is small, can squeeze A G C T on the other side 22:23 < fenn> er, uracil, sorry 22:23 < kanzure> huh? 22:23 < fenn> its the RNA version of thymine 22:24 < kanzure> right 22:24 < kanzure> but what does squeezing have to do with anything? 22:24 < fenn> you wouldnt be copying the template strand, just using it to crawl along 22:24 < fenn> the information comes from elsewhere 22:24 < fenn> so its not actualy a template, more like a scaffold 22:25 < fenn> so, when you add some arbitrary sequence opposite the scaffold strand, the bases dont really fit 22:26 < fenn> normally the "doesnt really fit" would activate the proofreading domain 22:27 < fenn> its kind of amazing we hav all this computer power and it isn't enough 22:28 < kanzure> for typing? :) 22:29 < fenn> for simulating protein, folding, and interactions between proteins 22:29 < kanzure> re: protein folding, see CASP, the protein folding prediction compettition 22:30 < fenn> say you have the sequence for pol-I, and the structure of DNA, you should be able to watch it go 22:30 < fenn> like that program Phun 22:31 < fenn> it doesnt have to correspond to reality 100%, just close enough that the same principles of design apply 22:31 * kanzure doubts we know protein engineering design principles :( 22:32 < fenn> well, we know the physical chemistry interactions between all the different amino acids 22:32 < kanzure> interesting 22:32 < kanzure> I just found a Nature review 'The descent of polymerization' 22:32 < kanzure> I clicky, and it's Andy 22:33 < kanzure> plus the last co-PI guy, Matt Levy 22:33 < kanzure> 'The in vitro selection of a ribozyme polymerase capable of catalyzing the faithful addition of up to 14 nucleotides to a series of noncovalently bound primers fills a niche in our understanding of the origins and evolution of life on Earth.' 22:35 < kanzure> A. Levskaya, A.A. Chevalier, J.J. Tabor, Z.B. Simpson, L.A. Lavery, M.Levy, E.A. Davidson, A.Scouras, A.D. Ellington, E.M. Marcotte, and C.A. Voigt (2005). Engineering Escherichia coli to see light. 22:36 < kanzure> http://ellingtonlab.org/mediawiki-1.10.0/index.php/Main_Page 22:36 < kanzure> guess I need to find a team that might do something related 22:36 < kanzure> Team ITS; Eric -- Emulsion selections of polymerase control regions (Near-term priorities) 22:36 < kanzure> Team Autogene; Eric -- Emulsion selections of polymerase enzymes Near-term priorities) 22:36 < kanzure> sounds about right? 22:36 < kanzure> haha, one guy. 22:36 < kanzure> "Team ITS. Eric." 22:38 < fenn> know any good protein visualization tools? 22:39 < kanzure> yes 22:39 < kanzure> no 22:39 < kanzure> http://heybryan.org/mediawiki/index.php/Computational_biology 22:39 < kanzure> http://heybryan.org/mediawiki/index.php/Comp_chem_linkdump 22:41 < kanzure> the #bioinformatics people will, if that comp bio page doesn't 22:41 < fenn> wonder if garlic can handle proteins well enough 22:41 < kanzure> eh? 22:44 < fenn> oh good, that's what it's for :) 22:44 < fenn> i knew i was onto something 22:45 < fenn> omg nice interface (not) 22:45 < kanzure> hm? 22:45 < kanzure> I'm going to email Eric 22:45 < kanzure> how do I sound, not stupid 22:45 < kanzure> guess I need to know what polymerase 'control regions' are 22:45 < kanzure> are they really control regions 22:48 < kanzure> ' 22:48 < kanzure> Whether or not genes are in an active or a repressed state in a cell depends on the relative effect of gene silencers and locus control regions (LCRs). Here, we suggest that these elements act as binary switches; the state that prevails (activated or expressed) probably depends on a competition between protein complex formation and the stability of the complexes formed at either of the two elements.' 22:48 < kanzure> oh, duh 22:49 < kanzure> so apparently 'enhancers' in transcription are related to the binding sites 22:53 < kanzure> http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2169052 Mutation at the Polymerase Active Site of Mouse DNA Polymerase δ Increases Genomic Instability and Accelerates Tumorigenesis 22:55 < kanzure> Mutation at the Polymerase Active Site of Mouse DNA Polymerase δ Increases Genomic Instability and Accelerates Tumorigenesis 5 bp core subunit enzyme? 22:55 < kanzure> erm 22:55 < kanzure> http://www.pingrysmartteam.com/RPo/RPo.htm 22:55 < kanzure> polymerase holoenzyme? 22:55 < fenn> enhancers are regions of DNA before a gene that attract transcription factors 22:55 < kanzure> ' In order to recognize a promoter to begin transcription, the 5-subunit core enzyme (α,α,β,β’,ω) must bind to one of various sigma (σ) factors; this form of the enzyme is called the holoenzyme. Each of the different σ factors recognize different promoter elements upstream of genes allowing the cell to respond to various environmental cues.' 22:55 < kanzure> ''Once holoenzyme binds the promoter, DNA downstream of this interaction is brought into the enzyme and melted to expose single-stranded DNA. This stage of transcription initiation is described here using a model of RNA polymerase-open-promoter-complex (RPo)' 22:55 < kanzure> so RNA polymerase-open-promoter-complex sounds important 22:56 < kanzure> now if only we had this for DNA polymerase 22:56 < kanzure> http://www.pnas.org/cgi/content/full/97/10/5095 DNA polymerase active site is highly mutable: Evolutionary consequences 22:57 < fenn> why do you want that for DNA polymerase? (it exists, look up 'ori site') 22:57 < fenn> (origin of replication) 22:57 < kanzure> oh 22:57 < kanzure> I already saw that paper 22:57 < kanzure> well 22:57 < kanzure> we want to find the complexes on the polymerase molecule 22:57 < kanzure> and then see if anybody has reverse engineered this and shown that this correlates to an *exact* subset of the entire gene for polymerase 22:58 < kanzure> "site-directed mutagenesis" 22:58 < fenn> reverse engineered the mapping between genome locations and functionality in the polymerase? 22:59 < fenn> or mapping between functionality and structural subunit in the polymerase? 22:59 < fenn> or something else? 22:59 < kanzure> the first 22:59 < kanzure> http://www.pubmedcentral.nih.gov.ezproxy.lib.utexas.edu/articlerender.fcgi?artid=452606 23:00 < kanzure> ' Exonuclease assays of the resulting mutant proteins indicate that the largest effects on exonuclease activity result from mutations in a group of carboxylate side chains (Asp355, Asp424 and Asp501) anchoring two divalent metal ions that are essential for exonuclease activity. ' 23:00 < fenn> ok that's easy. just knockout a particular nucleotide using "site directed mutagenesis" 23:00 < kanzure> here's what I'm thinking of doing 23:00 < fenn> an insertion will caluse a frame error 23:01 < kanzure> we'll take a molecular dynamics simulation package of polymerase 23:01 < kanzure> we'll then use this to just hack a way and add some extra functionality that we like 23:01 < fenn> also you can do bacterial transformation stuff, its like reading a tape 23:01 < kanzure> we'll then make some general simulations and predictions 23:01 < kanzure> and then we'll design selection experiments that would find that 23:02 < fenn> negatory 23:03 < kanzure> hm? 23:03 < fenn> molecular dynamics cant handle big roteins like polymerase, and on top of that, we can't hack in new functionality because we dont know how to do that 23:03 < kanzure> 40,000 atoms on a 'top of the line' consumer-market machine 23:04 < fenn> 40kD doing what? 23:04 < fenn> looking pretty? or actually functinoning 23:04 < kanzure> hrm, good question 23:04 < kanzure> also, we have Big Iron available at the lab 23:05 < fenn> with much red tape no doubt 23:05 < kanzure> apparently not 23:08 < kanzure> this one looks important - 'The Conserved Active Site Motif A of Escherichia coli DNA Polymerase I Is Highly Mutable' 23:10 < kanzure> haha 23:10 < kanzure> UmuC/DinB family (e.g. DNA polymerase ). Crystal struc- 23:10 < kanzure> tures of representative enzymes from the first four families 23:10 < kanzure> have been determined, revealing a common overall architec- 23:10 < kanzure> ture that has been likened to a human right hand, with fingers, 23:10 < kanzure> thumb, and palm subdomains (5–9). Although the structures of 23:10 < kanzure> the fingers and thumb subdomains vary considerably, the cat- 23:10 < kanzure> alytic palm subdomains are all superimposable (10, 11). The 23:11 < kanzure> acid residue. Essential roles of motif A in catalysis include 23:11 < kanzure> interaction with the incoming dNTP and coordination with two 23:11 < kanzure> divalent metal ions that are required for the polymerization 23:11 < kanzure> reaction (12–15). Motif A begins at an anti-parallel -strand 23:11 < kanzure> hm. 23:15 < fenn> here's a couple pics if it helps http://www.ncbi.nlm.nih.gov/Web/Newsltr/FallWinter02/structure.html 23:15 < fenn> another one, not pol-I though http://www.biochem.umd.edu/biochem/kahn/teach_res/Overheads/T7DNAP.jpg 23:19 < kanzure> btw, I'm not sure if looking at 3D visualizations 'helps' much 23:19 < kanzure> I mean, I'm still surprised that enzymes sometimes have 'knooks and cubbies' for other things to connect into 23:19 < kanzure> mainly because the mechanical probability of anything happening there seems small to me 23:19 < kanzure> won't it be mostly chemical interactions at that scale? so shape won't matter much? 23:22 < fenn> yes and no 23:23 < fenn> its the moore's law bug 23:23 < fenn> shape still matters even at the small molecule level 23:26 < fenn> ever used a microscope at 1000X zoom? bacteria are bouncing around all over the place from brownian motion, so their insides must be just whirring around at a tremendous rate 23:27 < kanzure> I like to imagine giant floating molecules of stuff from chem 101 interacting with other largely diffused molecules, and then radically changing shape 23:27 < kanzure> but I guess shape can be conserved in some reactions 23:27 < fenn> and the insides are not mostly water like they show in pictures, it's all really tightly packed in there.. i think its like 30% free water 23:29 < fenn> ah the power of the internet:, i found it: http://mgl.scripps.edu/people/goodsell/illustration/public 23:31 < kanzure> would it surprise you if I say I've seen that image before 23:31 < fenn> no 23:31 < fenn> i had it hanging on my wall for a couple years 23:33 < kanzure> big? 23:34 < fenn> scanned it from my mol-bio book and printed on a poster printer 23:35 < kanzure> who the hell named this thing 'motif A' 23:36 < kanzure> ooh 23:37 < kanzure> catalysis (35). Both the N- and C-terminal parts of motif A 23:37 < kanzure> tolerated a wide spectrum of substitutions. DNA polymerase 23:37 < kanzure> activity associated with single amino acid substitutions within 23:37 < kanzure> the N-terminal 5 amino acid residues was as high or higher 23:37 < kanzure> than that of the wild-type enzyme. These residues form part of 23:37 < kanzure> an anti-parallel -sheet structure that is believed to accommo- 23:37 < kanzure> date the triphosphate moiety of the incoming dNTP and may be 23:37 < kanzure> a potential target for engineering of pol I derivatives with 23:37 < kanzure> altered properties. In contrast, amino acid substitutions within 23:37 < kanzure> the C-terminal five residues tended to be associated with re- 23:37 < kanzure> duced activity. 23:37 < kanzure> "a potential target of engineering of pol I derivatives with altered properties" 23:37 < kanzure> uh, yeah 23:38 < fenn> http://www.inthenews.co.uk/news/science/scientists-launch-protein-folding-computer-game-$1222182.htm 23:38 < kanzure> yep 23:38 < kanzure> some 3D pathfinder or something 23:39 < kanzure> "because people know how to do it better" etc etc 23:39 < fenn> humans are supposed to be good at traveling salesman problems 23:40 < kanzure> has this been tested? 23:40 < kanzure> I see what you mean, but I'm wondering if we have some evidence 23:41 < kanzure> both DNA and RNA substrates. We conclude that isoleucine at 23:41 < kanzure> position 709 contributes to sugar discrimination by wild-type 23:41 < kanzure> pol I and that this function may promote conservation of the 23:41 < kanzure> wild-type motif A sequence. Based on analysis of a structural 23:41 < kanzure> fun stuff, I think this paper is a goldmine 23:43 < fenn> foldit looks really cool 23:50 < kanzure> I need a second search box in Opera (yes, I'm back to Opera ... for the mean time) 23:51 < kanzure> one that does Google Scholar directly 23:51 < kanzure> with my proxy stuff already done 23:56 < kanzure> Eric got back to me :(