--- Day changed Sun Feb 08 2009 00:03 -!- wrldpc [n=ben@c-98-217-190-207.hsd1.ma.comcast.net] has joined #hplusroadmap 00:12 -!- PeerInfinity [n=someone@216.36.180.162] has quit [] 00:26 -!- wrldpc [n=ben@c-98-217-190-207.hsd1.ma.comcast.net] has quit [] 00:31 -!- gene [n=chatzill@wireless-128-62-93-215.public.utexas.edu] has quit [Read error: 104 (Connection reset by peer)] 01:25 -!- wrldpc [n=ben@173.48.207.37] has joined #hplusroadmap 10:13 -!- samrose [n=samrose@c-24-11-214-181.hsd1.mi.comcast.net] has joined #hplusroadmap 10:49 < fenn> i think we've got ourselves a cad program: http://fennetic.net/pub/irc/heekscad_fillets.jpg 11:09 -!- PeerInfinity [n=someone@216.36.180.162] has joined #hplusroadmap 11:10 -!- wrldpc [n=ben@173.48.207.37] has quit [] 11:47 -!- samrose [n=samrose@c-24-11-214-181.hsd1.mi.comcast.net] has quit [Remote closed the connection] 12:07 < kanzure> hm, I wonder if it is worth the effort to get people to convert protocol-online.org into the clinical protocol XML format that I mentioned in the "recipe representation" thread 12:43 -!- elias` [n=me@unaffiliated/elias/x-342423] has quit [Read error: 110 (Connection timed out)] 13:00 < kanzure3_> So how am I supposed to know whether or not a "buffer" is a 'reagent' or a 'specimen' in the protocol? 13:00 < kanzure3_> http://www.mcb.uct.ac.za/pcrcond.htm for PCR. 13:01 < kanzure3_> I guess specimen is supposedly coming from 'patients', whereas reagents are required to make the protocol-magic happen. 13:02 < kanzure3_> (seeing as how CLP-ML comes from a clinical environment.) 13:02 < fenn> specimen changes whereas reagent stays the same? 13:02 < fenn> across different experiments 13:02 < kanzure3_> in the initiation? ok. I can go with that. 13:03 < kanzure3_> also, why publish a DTD, write up a paper about how awesome it is, and then fail to include a complete example XML file 13:04 < kanzure3_> haven't we come across this before in some other areas? at least the PSL people had an example (though a weird one) 13:04 < fenn> yeah you'd expect them to come up with hundreds of examples and test cases in the process of developing the standard, so why not include them 13:05 < fenn> i think the claimed reason is so they dont introduce "bugs" or test cases that dont mean what they ought to 13:05 < fenn> but that's lame 13:14 < kanzure3_> hrm, then there's another element - materials - which is at the same level as reagents, and the same level as specimenRequirements 13:15 < kanzure3_> the DTD: http://www.clinchem.org/content/vol0/issue2006/images/data/clinchem.2006.071449/DC1/clinchem.2006.071449-2.txt 13:15 < kanzure3_> so materials != instruments != specimens != reagents 13:15 < kanzure3_> what would count as a material? a test tube? or is that an instrument? 13:15 < kanzure3_> is water a material, or a reagent? 13:16 < kanzure3_> reagent = any chemical used in a laboratory test or experiment. 13:19 < kanzure3_> specimens: dna that you want to amplify, primers, .. 13:19 < kanzure3_> reagents: taq, DMSO, nucleotides 13:19 < kanzure3_> materials: buffer? test tubes? 13:20 < kanzure3_> instruments: open source gel box, thermocycler, straw :p 13:20 < kanzure3_> (this, kiddies, is why ontology sucks) 13:52 < kanzure3_> does buffer come in pcr kits these days? if so, that would imply that in this system, buffer should be a reagent, even though it seems more like a material. 13:53 < kanzure3_> (reagent has a subsection for 'kit' actually, but that's kinda cheating methinks) 14:15 -!- elias` [n=me@unaffiliated/elias/x-342423] has joined #hplusroadmap 15:16 < kanzure3_> http://heybryan.org/~bbishop/docs/protocols/pcr.xml 15:16 < kanzure3_> this does more damage than it does good. 15:20 < fenn> i dont get why buffer is one reagent 15:20 < fenn> "1.5mM Mg2+, usually some detergent, perhaps some gelatin or BSA" 15:20 < fenn> should be separate reagents no? 15:21 < kanzure3_> this ontology sucks. yes. /me fixes 15:21 < fenn> can you nest reagents? 15:22 < kanzure3_> 15:22 < kanzure3_> doesn't look like it. 15:22 < fenn> maybe that's what is for 15:23 < fenn> but i really think it should be nestable 15:23 < fenn> next point; is there a way to describe temperature changes in units? 15:24 < kanzure3_> even if there was, this would be something about the input/output to the thermocycler, no? 15:24 < fenn> no? 15:24 < kanzure3_> then what for? 15:24 < fenn> does a protocol describe the tools you use or just the process that 'stuff' goes through? 15:24 < kanzure3_> the process. 15:24 < fenn> i'd argue that they should be separate 15:24 < fenn> ok good 15:25 < fenn> so you talk about temperature, not about thermocycler settings 15:25 < kanzure3_> there should be a BOM, bill of technologies, and then some other things 15:25 < kanzure3_> hrm. 15:26 < fenn> of course there should be an explanation of how to set your thermocycler as well, but it shouldnt be the only way 15:27 < fenn> there's lots of human readable text explaining things that could be in code 15:27 < fenn> what's the purpose of this, btw 15:28 < kanzure3_> cat blah.bom > /dev/checkout 15:28 < fenn> so its just a list of materials? then why all the process stuff 15:28 < kanzure3_> instructions. 15:29 < kanzure3_> theoretically this should be a data format for protocol-online.org 15:29 < kanzure3_> instead of having all of those protocols written up in varying details 15:29 < kanzure3_> *varying detail 15:29 < fenn> how would protocol-online benefit from having semantic metadata? 15:29 < kanzure3_> well, this would be more for diybio people to use this information to see that they need to go get a certain kit, or whatever 15:30 < kanzure3_> codifying the information rather than just leaving it out in the mailing list, all raw and dead-like 15:30 < kanzure3_> *poke* 15:30 < fenn> hey is protocol-online.org just a bunch of links to other webpages? 15:30 < kanzure3_> yes :( 15:30 < fenn> fuck that 15:31 < kanzure3_> sucks huh 15:31 < fenn> they could at least periodically check to make sure the link isn't dead 15:31 < kanzure3_> most of the links are dead now 15:31 < fenn> well duh, that's what happens 15:31 < kanzure3_> internet archive favors .edu's or so it seems, so it's not entirely a lost cause 15:31 < kanzure3_> right 15:32 < fenn> how come you and me have to do this stuff.. 15:32 < fenn> tired of fixing this broken planet 15:34 < fenn> reagents should link to a dedicated page or xml file for that reagent 15:34 < fenn> unless it matters which supplier it comes from 15:35 < kanzure3_> ideally it shouldn't, but in practice I've heard it does. 15:36 < fenn> oh, and ng isnt a volume :) 15:36 < kanzure3_> maybe a better way to represent this information would be some way of separating reusable and one-time tools, materials, items, etc. 15:36 < kanzure3_> ng nanograms? 15:36 < fenn> ya 15:36 < kanzure3_> oh wait :) 15:36 < kanzure3_> blah 15:37 < fenn> what, you mean you dont reuse your pipette tips?! 15:37 < kanzure3_> http://www.mcb.uct.ac.za/pcrcond.htm <-- really I was just copying and pasting through this 15:37 < kanzure3_> heh 15:38 < kanzure3_> so, there's instructions and notes; list of reusable materials, tools, etc.; list of one-time use materials, tools, instruments 15:38 < kanzure3_> anything else? 15:39 < kanzure3_> (I'm effectively coming up with a brand new ontology, which is probably a bad idea, but it's not like anyone else has grown the balls to do things right the first time anyway.) 15:39 < fenn> how many turtles do you want? 15:40 < kanzure3_> turtles get their own files :) 15:40 < fenn> good 15:40 < kanzure3_> heh, that might be a fun python script to write. "please input an integer from 1 to 1000 for the number of turtles in this specification." 15:40 < fenn> microsoft ontology creation wizard 15:40 < kanzure3_> "I see you are writing a suicide note. Would you like assistance?" 15:43 < fenn> i think the most important part is the section, which seems to be rather neglected 15:43 < fenn> there should at least be some and