ucsf (very out of date)see:domsetup

|<*

*> WIKI page(s) here <**&here for protege+lisp interaction; >>os-mac,dell<<&wiki not open to public.

=The last-system-suggestions, (can modify by need/availability) machine summary.cl-*

=*Bug-tracking,available whenever you want it. Send me a bug& i will post it for you.

Write/

me on any of this. Right now these end up being more piece-meal notes to myself.

If a dialog is started, I will expand (any topic) here &/or in the wiki.] My home# is 312-666-3294

me.]

eg. Is there apc/mac that I could connect to via the network? got g4&updating it now..

Right now other machines have been more trouble than help; so far. [I'd like a usable OS.]

If I thought there was any chance we could actually use OSX I'd get a fully speced Intel 17"

• •  <even the past days links are only, on the new website> NO MORE POSTS HERE -GOTO NEW WEBSITE!!

  • 11 AM Tues. Mar 4, AM group meeting 

  • Mon Mar3, UMLS starts the use of V6  

  • weekend:  I'll be swtiching this page over to: http://domsetup.ucsf.edu:8080/mb/daily.htm

  • Fri f29:umls semantic-net webcast  ;biotop . . . ;Even though Ben will rework new umls data into old data schema, I wrote nlm asking for the new load scripts; &my pre-comment for today's talk is I'd like to see an easy way to get it into an instance/obj(or even triple)store, and very easy(rest-ful?)ws-api; After watching a few more screencasts, I decide to augment my question about the load files w/a request for the dvd if that is the only way to get the real ones.;I did find that mmsys.zip has the *mmsys.sh install files, but not sql scripts(yet). * . 

  • Thur, set up an admin list, since no wiki yet; You can open the latest onto via open-URL from here; Here is a standard html dump; I think an easier vers of a live website is coming as well. ;gs

  • Wednesday 2/27 at 2 pm. elig-rule-gram meeting   (pick: today or tomorrow on sysadmin);I was happy w/the call, it is good to see the use of SAGE, and pulling in some of the sumo and other ideas in as well; Samson is also right about it being time consuming to map to the eMR; I think a simple eMR and showing the match is more important, then the time consuming map back to whatever actually ends up being the flavorful eMR at that time (being able to(somewhat more easily)go to more than one would be a bonus)

  • tue:have a public apache page now, to link out to; mentioned protege-server that I originally had on the win-xp box &only recently got up again on the g4;want umls up;.. A few pprjs you should be able to get to now.;I might make a homepage, so I can get a 6month trial of racer, which could replace ocelot/gkb, be seen via Protege, &even has a quick web display.;tried&got cl- http -load on deb box, will wrk on pick/open a port to run;but had to reset rctbank inetd.conf to turn off services;[darwin&deb are running cl-http on 8080, but no pages are assigned yet] &should prep4call

• • /other;

  • mon o29,,in_ex-clusion, is umls only networked or is it on a local machine?

  • weekend,cli,nh,,ebp,ncboTools,

  • fri-osx26,caBIG_Rules_Engine_WP.doc,ht,..,,,,a,,

  • thur o25,,,lps;,,fz-sw/dl book

  • wed o24,,,,;tc,,*,,ccrMatch.cl,would like to know more about(how we use)umls/snomed&some of Ben's related work;lx,got: NCBOsoftwareCurrentStateReport.doc,but I can't distribute it;obd,,a,

  • tue o23,g,,,cusp.,,,,still reading, go 1st Qfrom Ben for repeat of machine info/still want2see a lnx run

  • mon o22, load of:ccr.cl;>,r,,Bug-tracking-&people_pages, reading

• weekend;I'd like to get a better idea of what is going on; Might write smi? Nothing from Dr.Sun or Ben yet.

  • fri o19,problem w/ebay sun parts /need support.(go2LH maybe meet ben?)/restart machines*;,ccrFacts,,

  • thu o18,ccr xml importqry,p55;pomr,;(reading to assure path will work well w/smi&other work)

  • wedo17, some umn ccr info, incl gateway-schema,they use xPath Queries; smi feedback;;

  • tueo16umls,CCR site up this afternoonall MS/win files so trying wine*installs,but no loginmono might allow looking through win-schema files?; Will get more ccr-links soon.;it's an xml schema;also have more on an appliMatch test-case

  • mon o15,look@umn, then smi a bit/snapshot of my notes/etc on lnx box;

  • weekend,,LispWorks on lnx-dell, while work/ing w/cl-http

  • fri o12, from a doc 1link down;ccr repl lnk also still down;vpn allows hitting web but stops vpn;looking@opeing ports(if blocked beyond this box, I hope that has been passed on to me, by ben);reading more/finding places4improvements

  • thur oct11, got appli-case, another bad ePCRN site;look@firewall(might get external vnc to lnx via mac letting me in 1st,2much security still); look@admin &other docs; reading /will work in ~new ideas into test code fri

  • wedn o10,, LH meeting in the morning;helped further understand applimatch incoming from re; asked4ePCRN site;i'll be offline in the evening-till-afternoon

  • tue 09, got franz lic, get files from ben on arches?/trying to do it myself; maybe find out what is there&how to run it if lucky; also look at my own code again.;number questions to ben is ~helping some of them not be ignored.;I put fromBen.rtfd &./ben excerpts on rctbank;

• mon o8, write franz for our devel.lic for linux&arches[going w/acl8&express for now]; updated kepes to opensolaris [I'll back up some files just in case.;maybe over a period of time, till arches is fully going (as a backup); as no1 replied about kepes] I wanted to start a section on 'goals' which will work in well w/today's meeting(on an action plan).;Mv-ed enough files on kepes to other machines to use vi again.

  • weekend.;I've got a number of open questions; I'd like to reduce this list...! (ie.kepes, acl, arches)..;Found stanford epoch site<=are we connected to this?. ;newest cl-http port compiles w/acl8 on arches&dell ;I'd still like the other code ..

  • fri o5, lnx install(s), on dell&mac, I'll use the Debian-family, but can switch. [Dell would go from having Debian in vmware, to being dual boot; I was going to update the g3 first, as a trial.]Ended up w/a Debian on the dell, but no where else right now; I'll finish updating it today(apt-get);Got mysql&a try@cl-http &vnc;=What is kepes going to be used for/could it become a test machine if backed up?

  • thur oct4,I will try to compile GFP on a non-solaris OS, though I only have McCLIM; =Could I get a devel.lic for my laptop? We have a site-lic =(actually Why do we have that, do other labs use it as well)? ; Even for GKB we can use open-solaris on any dell(like)box, if not linux. [Updated sun(blade upgrade)quote is $13.1K/machine, so that is a no go; but look into ebay parts if you think it will help, &is cheaper than a starter lnx-box] -I'll do the 1st lnx install today.;From sys.htm, here is the dept suggested pdf.

  • wedo3,Get on IM and fix email. [Email looks to sorted; I have back emails if anyone is interested.] Sent email in my rctbank home-dir. Hope to stay away from machines&work on code. Here is a summary. Blades come w/lots of possible other service/etc options.;I might write a gfp interface to allegro-store/graph; trying to compile okbcfor protege interactivity.*

  • tue o2, pick up cables, re-connect-PC, ask for bugs to enter again; machine updates..Arches is still live; contact me for login email info via skype. [Linux would be easier to maintain, but can't get it to boot on these chips anyway.]kepes has poor partitioning (no space errs)&OS too old for ACL, I'd like to update it. ;from records, no1is actively using it; does it have a strategic plan?; umlisp on kepes, but no acl on arches yet

  • mono1,=what are the needs/time-constraints for the (sun/changes) for a new machine decision?. More-diss. =updated ucsf email, what freq are you getting?

  • weekend, reading dissertation(1/2) [amongst the long list of Q's for Ben, is the background research affecting design-decisions]&will he 'use' IM now?; simple-fz-rule,code

  • fri s28,=Was arches ever on the netwrk@this location? Looks like maybe no. I am not ok w/spending more than 1k on these old machines.(even that I think is a waste of$s)vs.a new option machine; Arches online; no extra cable;so pc offline, so ben.. can play w/arches [I have protege&cmucl on it]have acl81 but =where is it's devel.lic?

  • thur s27, pick up cd-r;try networking&if not a new OS; amia,p

  • wed s26, we should all have skype soon; If I'm offline, shoot me an email&I'll start it up.;ccr;rule-testing;comments on the applimatch study<-

  • tue s25, meeting ;then/call sun?&try network quickly [I'd rather run OpenSolaris or Linux on them] {hope4machine info from Chris then?;&Ben to talk about the code};Would like to get back2looking@CCR&related.;f-o

  • mon s24, look over present code for what might be used; how are snomed&other grouping presently becoming CLOS classes? &prot/fztry0(not as nice as rule using lisa.sf.net) {sun feedback from ben}

  • weekend, look@rules [needed(machine)changes];wrkOn DfCls that is in OcelotEd,would be great2have some1who would chat; still made a simple rule

  • fri s21, re/fixing mac; contacted chris/sun again; try2get arches to boot from cd2repair it;[ps sun still sells ultra's but not sun-blades?]Got info, not worth upgrade, * Arches has an os upgrade that let it get outbound networking, &protege etc installed.

  • thur s20, LH meeting 2pm, called sun again, got os-x on g3(trailbank.servehttp.com), Found out my ucsf email wasn't getting to DrSim/had a good meeting; Will spec out bug-server quickly, then prep for Tue. [Machinewise:get sun#/prob-ticket in;See about HW;but PC&Mac might also help for now.]

  • wedn s19, LH meeting postponed; I'm realizing I might not get a history &/or talks of design; much less an intro; So my work (direction) might not have as tight a fit as everyone might like @1st. /vmware-deb on pc;*updated*Wiki page(s) *

  • tue s18,(*ON* wiki) [hardware is at a stand-still; I'm waiting for sun &/or a new purchase.]

  • mon s19, [checked out techstore/etc] backup G3's files &reformat(as it was2be tossed;It can't be updated/too old); (work on arches OS upgrade; it is stuck); get cable/etc(soon, but won't matter till I get more work from sun again); Best Dev env is rctbank or the PC; If not our home systems.

  • weekend+, went through all the security trainings

  • fri s14, I'd like to work on protege&talk w/Ben about the rule-work; but he has been offline >1wk(most of my time here);&I'll probably have to go back in to struggle w/the old/broken? machine(s). [~1st week]

  • Talked w/Chris@sun-support, gave him some info/he is tracking down the rest;Wrote Emily about cable/etc purchase.

  • thru s13, (long)Orientation, then try to setup arches, which isn't networked yet. [SUN is very un-friendly.]We are down on supplies, not even cat5/rj for arches;(given turn-around times for Q&A, I'm going to have to just start doing things;&look into purchasing..[arches needs OS-upgrade, maybe another HDfix, sun doing some admin; (&/or replacing it is still an option)]

  • wed s12, clean up bug proj;got some info from simona; ucb-library-day;

  • tue s11, I can invite the sun-rep this week, but it will be embarassing as I can't log into most of the machines, &so don't have a full idea of what we have. Ben has ben unavailable even via email. I'm finding load-lisp.htm to be unusable.[I can't get log in, nor display new files.] We should have one login, w/nfs mounted home-dirs, and be able to easily get to each machine. I'd rather use a PC now. A bright note: you can open a shared Protege session from the office; So I'lltry this; Which also can be used for editing the kb. [I can add projects thatcan be selected from remotely. *SUMMARY*]

  • mon s10, goto LH, get id, work on kepes&pc more; try to get other logins &Bens contact info; I'm leaving the PC on for use w/vnc now.

  • weekend; will installbug-tracking to try it; might put my mockuponline as well mon.;=Does anyone rememberdesign-a-trial?

  • fri s7, more mail/vpn cfg; mailed sun repr again; would like Ben (phone/person)contact; If we use src-cntrlbug-trac/wiki might be better than bug-zilla;CL-darcs

  • thur s6, get ncbo wiki login;[ncbbo&victoria use bugzilla, so we could try it as well]but I'd still rather make one that is closer to the system that we are working on; getting more power&practice.[Also the idea of making explicit dependencies(even in daily work)is not a bad one.] =What did people think of eligWriter?; I have email&vpn now.

  • wedn s5; 1st(partial)group meeting; put vnc on pc(works)&on sun(but blocked);=Need a Ben meeting. Will look@bug-zilla,collabtrac/wiki; If I get the present bug list I can see how well my idea for this works.

  • tue sep4, agraph/racerVtb;would like to talk w/Ben about rule-work(as I might start w/this)&get a general overview-soon.

  • labor-day-wkend, (more from thu/fri) ..I still need an overview & some idea when I can get it. I'm hoping for more on the sets of problems that need to be solved; and how the present attempts have evolved to try to meet these changing needs. Then more on what the open areas are, & the places where I can help.

  • fri aug30, ID appt, try to meet w/Ben (when ever he is available); priced-a-mac;(vs. sun)

  • Ben will check w/ENS about VPN access again; maybe w/VNC.

  • =Has anyone looked@a protege eligibility&umls/etc tabs?

  • thur aug29, schedule ID&fingerprints; mention o-device*+;&something like: http://sigswo.org/papers/SIG-SWO-A503/SIG-SWO-A503-03.pdf (though I still have to say I'd like to stay away from xml/rdf formats until it is absolutely necessary) [noticed no property hierarchy].

  • Some points to look over before(&after) I get in town:

  • UMLS

  • I have: (from Ben)

  • UMLS is the main standardized medical vocabulary that we've used up to this point for computably representing CT components. NLM's claim is that UMLS subsumes SNOMED CT, but that seems to be controversial to some. We aren't necessarily tied to using UMLS over the long term, but we probably will continue to use something that resembles it, so I think becoming familiar with it will be helpful. A good starting point for looking at it would be http://www.nlm.nih.gov/research/umls/. The main documentation page is http://www.nlm.nih.gov/research/umls/documentation.html. I think the main thing to get a sense of at first is the overall organization of the UMLS data tables and the interrelationships between the three main groups of tables, META, LEX and NET.

    • For a Lisp perspective on UMLS, please look at Kevin Rosenberg's UMLisp at http://www.cliki.net/umlisp as one way of modeling UMLS in CLOS.

    • <QUESTION: I can't find /space/UMLS >The UMLS sources are on rctbank in the /space/UMLS directory. Sad to say, the latest version of UMLS that's installed now is 2004AB - this is one of the main things we need to upgrade. If you find that you need to get a username and password to access some code or utils from the NLM site, please let me know and I'll put you in touch with someone there who can arrange that. Btw, the UMLS sources take up several GB when installed.

    • At some point you may also want to subscribe to UMLS list, instructions at [which I have; Also I've loaded the code, but not the data.]

  • CL-HTTP (you can start a port on rctbank to work on this) [ask Ben for best current pointers on this]

  • It sounded like Ben had a good start on this.

  • clinical trials [Simona, can you send some introductory pointers to Michael? Perhaps he’d like to look at OCRe too.] [review RCT Schema, BaT and Presenter code]

  • I can look at provided papers, and owl file [also: http://rctbank.ucsf.edu/ontology/outline/index.htm]

  • From Simona:

  • In terms of introduction to clinical trials, I find the tutorial on the NCI web site quite nice. The drawback is that it is very cancer-focused. However, most concepts are generic and apply to all clinical trials (phase, eligibility, randomization, etc.):

    • http://www.cancer.gov/clinicaltrials/resources/basicworkbook/page1

    • I also found two short summaries that may be useful.

    • http://nextgenmd.org/vol1-4/clinicaltrialsv1i4.html

    • http://nextgenmd.org/vol1-5/clinicaltrialsv1i5.html

    • Then there is our project web site and you may want to start by looking at the intro page and follow the references. The publication page is not up to date, so I am sending you some recent papers (actually, two papers and a poster description). Finally, this is a recent paper by Ida published on PLoS Medicine:

    • http://medicine.plosjournals.org/perlserv/?request=get-document&doi=10.1371%2Fjournal.pmed.0020365&ct=1

    • <amia2003_1076.pdf><simtrialbank.pdf> <sdarticle-1.pdf>-(mb:I think I once got the Smalltalk code from Chandrasekeran's lab, but have since lost it)

  • our various bits of software [upgrade our systems as needed (Ben will let you know)]

  • I will start to look around on rctbank.

I have IM/Voice/Video via: skype://michael_bobak also that name for Yahoo-chat, and my gmail account for gTalk.

This started as "Some points to look over before I get in town: " but I might keep a short running log as well. 

=I'm not getting much interaction; Which I'm ok w/if I can make my own decisions; but if I'm asked to do things where cooperation is called for; I expect it.  [Maybe this will work out, So I'm moving some warning from above, here to the bottom:I have some questions, &serious issues below. <Weekly Summary Here> [lost email is on rctbank in ~michael/sent*] [Write me to set up a quick skype session on any of this. Rigt now these end up being more piece-meal notes to myself. If a dialog is started, I will expand (any topic) here &/or in the wiki.] My home# is 415-282-2682; I'd like Ben's#|IM =I've spent an all too large amount of time on the machines; It doesn't have to be like this; I'd like to change this.Just for the record, ben's not helping me get started is noticeably destructive to productivity. [I've asked for in person meetings, or at least by phone, then at least via chat (as email is a waste of time, as questions aren't getting answered)] Every day there is something I could do faster or at all if I got help from ben, but I'm not (for the most part).