- Microbe - mailman3.cels.anl.gov

Re: [Microbe] [Kbase-devel] Comments from users at GLBRC retreat
by Thomason, James 24 May '13

24 May '13

Awesome. As tutorials are updated, please conform to the structural requirements that iris needs to parse them out, then I can link 'em all into there as well. In iris, type 'tutorial list' to get the list of tutorials, and then click on "How to create an iris tutorial" and that'll talk you through the structure. It's basically structured as the pages currently are, just tacking on a 'id="tutorial" to the container, and wrapping individual steps in div tags. So modifications should be minimal. But once that's in place, any pages that follow the structure can be linked into iris and runnable there as well. -- -Jim Thomason... Computational Science Developer @ The Ware Lab, a USDA-ARS Laboratory at Cold Spring Harbor Laboratory http://www.warelab.org/ http://www.cshl.edu/ On May 24, 2013, at 4:33 PM, Elizabeth M. Glass <marland(a)mcs.anl.gov<mailto:[email protected]>> wrote: Thanks Chris! Just as FYI - 2. Good comment and easily fixed. 3. Lynn and I are reviewing the command-line help and a command summary is an item we would like to see added along with a usage example for each, as most end users will still have trouble deciphering how to use it. We are reorging the tutorials site as we speak - it is too confusing with it all mixed together. Elizabeth ----- Original Message ----- From: Christopher Henry <chrisshenry(a)gmail.com<mailto:[email protected]>> To: kbase devel <kbase-devel(a)lists.kbase.us<mailto:[email protected]>>, outreach(a)kbase.us<mailto:[email protected]> Outreach <outreach(a)kbase.us<mailto:[email protected]>>, KBase Microbes List <microbe(a)lists.kbase.us<mailto:[email protected]>> Cc: Doreen Ware <ware(a)cshl.edu<mailto:[email protected]>> Sent: Fri, 24 May 2013 16:26:59 -0500 (CDT) Subject: [Microbe] Comments from users at GLBRC retreat Hi all, I'm going to do a brain dump of comments from users at the GLBRC retreat. I figure everyone should see these. Doreen, can you add to this list? 1.) David Benton said we should do everything in "awk", but I think he was kidding :) 2.) Yury Bukhman said we shouldn't refer to the apps in KBase labs as prototypes, because it scares everyone away from using them. Currently, the KBase slide stack calls them prototypes. 3.) Audry Gasch gave some awesome input. She was incredibly supporting and helpful during my demo. Here's a few of her comments: -She found the "usage" of the current commands confusing (but she did like that it was there). She thinks the command "usage" should always start with a high level sentence explaining what the command do. -She found the documentation on the KBase website overwhelming. In particular, she didn't like how the "Command line" documentation and the "Web API" documentation was all mixed together. She would have preferred to see those separated out. But also, she would like to see the same sort of "command links" that you have in IRIS in www.kbase.us<http://www.kbase.us> as well, where you can go to "Modeling Scripts", then look for your command of interest, and click on it to see detailed descriptions. She also said the command documentation was also too low level. It didn't provide high-level descriptions of what each command was fundamentally meant to do. -She wanted ALL the documentation to be available somehow online within IRIS. For example, could she type "annnotate_genome --docs" and see the complete documentation of the command with all detail thats on www.kbase.us<http://www.kbase.us> instead of just the command usage -She loved the idea of the tutorials being built into IRIS, but there needs to be better descriptions of how to use them there. -Audrey said she won't use anything in KBase unless she has a good understanding of the fundamental methods the commands use in the background. In the other words, our command documentation should contain links to the papers describing their methods whenever possible, or detailed descriptions of the entire method when no paper exists (e.g. she wants to know that RAST is the annotation technology behind our annotation commands, and Model SEED is the tech behind our modeling commands). 4.) After I ran "fastsa-to-genome" and "annotate-genome", many users asked if they could now go and see their genome in the Workspace Browser and in the Genome Browser. 5.) Steven Slater said he wants his agrobacteria community to set up a workspace with snapshots of all the agrobacteria genomes in KBase, then he wanted his community to be able to annotate those genomes there and share them as the best available agrobacteria annotations available. He also ultimately wanted to see these get integrated into the automatically propagated annotations. 6.) Generally, there was the question is how JGI data would be available in KBase… do we replicate it, or is it "just there" Chris _______________________________________________ Microbe mailing list Microbe(a)lists.kbase.us<mailto:[email protected]> https://lists.kbase.us/mailman/listinfo/microbe _______________________________________________ Kbase-devel mailing list Kbase-devel(a)lists.kbase.us https://lists.kbase.us/mailman/listinfo/kbase-devel

1 0

Re: [Microbe] [Kbase-devel] Comments from users at GLBRC retreat
by Land, Miriam L. 24 May '13

24 May '13

On 5/24/13 5:26 PM, "Christopher Henry" <chrisshenry(a)gmail.com> wrote: > > >3.) Audry Gasch gave some awesome input. She was incredibly supporting >and helpful during my demo. Here's a few of her comments: > -She found the "usage" of the current commands confusing (but she did >like that it was there). She thinks the command "usage" should always >start with a high level sentence explaining what the command do. Ditto! Miriam

1 0

Re: [Microbe] Comments from users at GLBRC retreat
by Elizabeth M. Glass 24 May '13

24 May '13

Thanks Chris! Just as FYI - 2. Good comment and easily fixed. 3. Lynn and I are reviewing the command-line help and a command summary is an item we would like to see added along with a usage example for each, as most end users will still have trouble deciphering how to use it. We are reorging the tutorials site as we speak - it is too confusing with it all mixed together. Elizabeth ----- Original Message ----- From: Christopher Henry <chrisshenry(a)gmail.com> To: kbase devel <kbase-devel(a)lists.kbase.us>, outreach(a)kbase.us Outreach <outreach(a)kbase.us>, KBase Microbes List <microbe(a)lists.kbase.us> Cc: Doreen Ware <ware(a)cshl.edu> Sent: Fri, 24 May 2013 16:26:59 -0500 (CDT) Subject: [Microbe] Comments from users at GLBRC retreat Hi all, I'm going to do a brain dump of comments from users at the GLBRC retreat. I figure everyone should see these. Doreen, can you add to this list? 1.) David Benton said we should do everything in "awk", but I think he was kidding :) 2.) Yury Bukhman said we shouldn't refer to the apps in KBase labs as prototypes, because it scares everyone away from using them. Currently, the KBase slide stack calls them prototypes. 3.) Audry Gasch gave some awesome input. She was incredibly supporting and helpful during my demo. Here's a few of her comments: -She found the "usage" of the current commands confusing (but she did like that it was there). She thinks the command "usage" should always start with a high level sentence explaining what the command do. -She found the documentation on the KBase website overwhelming. In particular, she didn't like how the "Command line" documentation and the "Web API" documentation was all mixed together. She would have preferred to see those separated out. But also, she would like to see the same sort of "command links" that you have in IRIS in www.kbase.us as well, where you can go to "Modeling Scripts", then look for your command of interest, and click on it to see detailed descriptions. She also said the command documentation was also too low level. It didn't provide high-level descriptions of what each command was fundamentally meant to do. -She wanted ALL the documentation to be available somehow online within IRIS. For example, could she type "annnotate_genome --docs" and see the complete documentation of the command with all detail thats on www.kbase.us instead of just the command usage -She loved the idea of the tutorials being built into IRIS, but there needs to be better descriptions of how to use them there. -Audrey said she won't use anything in KBase unless she has a good understanding of the fundamental methods the commands use in the background. In the other words, our command documentation should contain links to the papers describing their methods whenever possible, or detailed descriptions of the entire method when no paper exists (e.g. she wants to know that RAST is the annotation technology behind our annotation commands, and Model SEED is the tech behind our modeling commands). 4.) After I ran "fastsa-to-genome" and "annotate-genome", many users asked if they could now go and see their genome in the Workspace Browser and in the Genome Browser. 5.) Steven Slater said he wants his agrobacteria community to set up a workspace with snapshots of all the agrobacteria genomes in KBase, then he wanted his community to be able to annotate those genomes there and share them as the best available agrobacteria annotations available. He also ultimately wanted to see these get integrated into the automatically propagated annotations. 6.) Generally, there was the question is how JGI data would be available in KBase… do we replicate it, or is it "just there" Chris _______________________________________________ Microbe mailing list Microbe(a)lists.kbase.us https://lists.kbase.us/mailman/listinfo/microbe

1 0

Comments from users at GLBRC retreat
by Christopher Henry 24 May '13

24 May '13

Hi all, I'm going to do a brain dump of comments from users at the GLBRC retreat. I figure everyone should see these. Doreen, can you add to this list? 1.) David Benton said we should do everything in "awk", but I think he was kidding :) 2.) Yury Bukhman said we shouldn't refer to the apps in KBase labs as prototypes, because it scares everyone away from using them. Currently, the KBase slide stack calls them prototypes. 3.) Audry Gasch gave some awesome input. She was incredibly supporting and helpful during my demo. Here's a few of her comments: -She found the "usage" of the current commands confusing (but she did like that it was there). She thinks the command "usage" should always start with a high level sentence explaining what the command do. -She found the documentation on the KBase website overwhelming. In particular, she didn't like how the "Command line" documentation and the "Web API" documentation was all mixed together. She would have preferred to see those separated out. But also, she would like to see the same sort of "command links" that you have in IRIS in www.kbase.us as well, where you can go to "Modeling Scripts", then look for your command of interest, and click on it to see detailed descriptions. She also said the command documentation was also too low level. It didn't provide high-level descriptions of what each command was fundamentally meant to do. -She wanted ALL the documentation to be available somehow online within IRIS. For example, could she type "annnotate_genome --docs" and see the complete documentation of the command with all detail thats on www.kbase.us instead of just the command usage -She loved the idea of the tutorials being built into IRIS, but there needs to be better descriptions of how to use them there. -Audrey said she won't use anything in KBase unless she has a good understanding of the fundamental methods the commands use in the background. In the other words, our command documentation should contain links to the papers describing their methods whenever possible, or detailed descriptions of the entire method when no paper exists (e.g. she wants to know that RAST is the annotation technology behind our annotation commands, and Model SEED is the tech behind our modeling commands). 4.) After I ran "fastsa-to-genome" and "annotate-genome", many users asked if they could now go and see their genome in the Workspace Browser and in the Genome Browser. 5.) Steven Slater said he wants his agrobacteria community to set up a workspace with snapshots of all the agrobacteria genomes in KBase, then he wanted his community to be able to annotate those genomes there and share them as the best available agrobacteria annotations available. He also ultimately wanted to see these get integrated into the automatically propagated annotations. 6.) Generally, there was the question is how JGI data would be available in KBase… do we replicate it, or is it "just there" Chris

1 0

Re: [Microbe] what are the proper flags for gapfilling a model?
by Christopher Henry 23 May '13

23 May '13

Paramvir, I actually removed the "no submit" thing. I removed cluster services entirely from the FBA. Instead, a job is created and placed in the workspace, and a separate generic scheduler script runs that job. My thought is that the scheduler should handle submission to nersc when we're ready to do that. It shouldn't be done directly by the FBA service. This gives us maximum flexibility in handling the FBA jobs in KBase. Chris On May 23, 2013, at 2:18 PM, Paramvir Dehal <psdehal(a)lbl.gov> wrote: > Matt, did anyone answer your questions? > > The --nosubmit is still necessary as far as I know. > > > On Tue, May 21, 2013 at 9:01 AM, Matt DeJongh <dejongh(a)hope.edu> wrote: > I remember that there is an undocuments --nosubmit or something like that ... is it still necessary? Anything else? Thanks, > > Matt > > -- > Matt DeJongh, Ph.D. > Computer Science Department > Hope College, Holland, MI 49423 > (616) 395-7429 > > _______________________________________________ > Microbe mailing list > Microbe(a)lists.kbase.us > https://lists.kbase.us/mailman/listinfo/microbe > > > _______________________________________________ > Microbe mailing list > Microbe(a)lists.kbase.us > https://lists.kbase.us/mailman/listinfo/microbe

1 0

Re: [Microbe] what are the proper flags for gapfilling a model?
by Paramvir Dehal 23 May '13

23 May '13

Matt, did anyone answer your questions? The --nosubmit is still necessary as far as I know. On Tue, May 21, 2013 at 9:01 AM, Matt DeJongh <dejongh(a)hope.edu> wrote: > I remember that there is an undocuments --nosubmit or something like that > ... is it still necessary? Anything else? Thanks, > > Matt > > -- > Matt DeJongh, Ph.D. > Computer Science Department > Hope College, Holland, MI 49423 > (616) 395-7429 > > _______________________________________________ > Microbe mailing list > Microbe(a)lists.kbase.us > https://lists.kbase.us/mailman/listinfo/microbe > >

1 0

Re: [Microbe] workspace browser - choose workspace service?
by Neal Conrad 22 May '13

22 May '13

Hi Matt, Do you mean use a different workspace via something like "/browser.html/workspace/<workspace_name>"? If so, no, there's no url routing right now. I've been learning the ws browser code and working on some new features, so I'll append this to the list. Thanks, Neal On May 22, 2013, at 8:05 AM, Matt DeJongh wrote: > Is there a way to configure the workspace browser to use a different workspace service by specifying the URL? > > -- > Matt DeJongh, Ph.D. > Computer Science Department > Hope College, Holland, MI 49423 > (616) 395-7429 > _______________________________________________ > Microbe mailing list > Microbe(a)lists.kbase.us > https://lists.kbase.us/mailman/listinfo/microbe

1 0

workspace browser - choose workspace service?
by Matt DeJongh 22 May '13

22 May '13

Is there a way to configure the workspace browser to use a different workspace service by specifying the URL? -- Matt DeJongh, Ph.D. Computer Science Department Hope College, Holland, MI 49423 (616) 395-7429

1 0

what are the proper flags for gapfilling a model?
by Matt DeJongh 21 May '13

21 May '13

I remember that there is an undocuments --nosubmit or something like that ... is it still necessary? Anything else? Thanks, Matt -- Matt DeJongh, Ph.D. Computer Science Department Hope College, Holland, MI 49423 (616) 395-7429

1 0

3 month project objectives
by Paramvir Dehal 20 May '13

20 May '13

Hi guys, Let's try and do a group edit in the google doc: https://docs.google.com/a/lbl.gov/document/d/17IF8zbzk77mdRlvkdFEKESe9zQDqL… * Please indicate which lettered tasks you would like to lead by writing your name next to it in bold. Indicate which tasks you would like be a team member of by writing your name next to it. If you think a task is already done, a mistake, or should be removed from the list, add a comment with your reason why it should be removed. If you think something is critical/very high priority, indicate that in the comments as well. If there are tasks missing, add them to the end of the list. Do not mess up the numbering!!!!*

1 0