ICIS Developers Workshop Australia 2010

From ICISWiki

Jump to: navigation, search

Contents

HOSTING

Hosted by The International Centre for Plant Breeding Education and Research (ICPBER)

photo credits: ching habito
Venue: University Club, University of Western Australia
35 Stirling Highway, Crawley WA 6009

REGISTRATION

Please have the following form filled-out and returned to Sarah Mawson not later than 1 February 2010: Download registration form


Expenses for this workshop There is a non-profit making fee of AUD$400.00 (+ $40 GST) for this workshop. This will cover hire of the facilities and all catering during the days of the workshop (morning and afternoon teas, plus lunches and 1 evening barbecue). All expenses such as accommodation, transport, and other evening meals, plus social activities on Saturday, will be borne by the participants.

How to get to the Workshop venue: The workshop will be held at the University Club, University of Western Australia 35 Stirling Highway, Crawley WA 6009. It is located within a brisk 20 minute walk , or short bus ride, from the Kings Park Motel, Subiaco and Sullivans Hotel, Perth (See accommodation).

Accommodations: Workshop participants will be expected to arrange for their own accommodations.

Recommended hotel(s):

More luxurious hotels can be found, but further away from the University.

Food: Lunches and morning and afternoon teas will be provided during the days of the workshop within the fee. There will also be a workshop evening barbecue, which at present is included in your fee, but for which we are seeking sponsorship.

For special dietary needs inquiry, please email Sarah Mawson  and every effort will be made to accommodate those with special dietary needs.

There are many restaurants in Perth and Subiaco

City Information: Please visit this site for useful information about Perth.

Questions? Please direct any question to Sarah Mawson .

AGENDA

Please place links to Powerpoints, videos, other presentation materials in the DISCUSSION section. Thanks.

Date Time Activity Facilitator
Reporter

Day 1 -

Tuesday,
2 Mar 2010

0830
 Registration
0900
 Welcome Address - Willie Erskine

Facilities Orientation
Group Photo

Workshop Objectives and Agenda - Thomas Metz
1030
 Coffee break

Activity Reports of ICIS Partners (10 minutes each)

IRRI (Thomas)
CIMMYT (Arllet & Juan Carlos)
AAFC (Shawn Yates - reported by Arllet)
BAYER (Ed)
CBWA (Cameron)

 


New Members Key Interests in ICIS (10 minutes each)

UA (Howard Eagles)
GQNPC-IRRI (Roslen)
C4-IRRI (Richard)

1230
 Lunch

New Members Key Interests in ICIS (continued)

ICRISAT (Trushar c/o Arllet)
GCP/MBP (Graham)

1430
 Coffee break

***Howard Eagles public lecture***

 

Day 2 -

Wednesday,
3 Mar 2010

8:30

Starting to use ICIS:

Marketing ICIS through the new ICIS Customer Webpage (Warren)
New IRIS Installer (Ching)
GCPwizard + File Conversion Wizard: Migrating into ICIS the easy way (Warren)

(F) Richard

(R) Mylah

10:30
 Coffee break

Customizing ICIS:

Development of Generic GRIMS (Juan Carlos)
Linking GRIMS with ICIS for Seed Multiplication (Beth)
Splitting Setgen into two use cases? (Ruaraidh)
Development of the Configurable Workflow (Chengzi)

(F) Richard

(R) Mylah

12:30
 Lunch

Customizing ICIS (continued):   

     Supporting User Workflows through the ICIS Lite (Warren)

Entering data in ICIS Part 1: Genealogy Management

Managing documents using SetGen (Grace)
Admin utility for handling persons, institutes, addresses, and locations data (Grace)

(F) Ruaraidh

(R) Beth


15:00 		Coffee break

Entering data in ICIS Part 1: Genealogy Management (continued):

SetGen and GMSsrch UI changes (Ching)
ICIS Schema 5.6 (Grace)

(F) Ruaraidh

(R) Beth

Day 3 -

Thursday,
4 Mar 2010

8:30

Entering data in ICIS Part 1: Genealogy Management (continued):

ICISwbk-GMS: Managing germplasms with ease and comfort (Warren)

Entering data in ICIS Part 2: Inventory Management

ICISwbk-IMS: Tracking everything through barcoding (Warren)

(F) Arllet

(R) William


Coffee break

Entering data in ICIS Part 3: Data Management

ICISwbk-DMS (Warren)
ICIS Mobile: Taking ICIS out into the field (William)

(F) Arllet

(R) William/Arllet

12:30
 Lunch

Entering data in ICIS Part 4: Gene Management

ICISwbk-GEMS (Warren)
MBP Plans - ICIS 6 Schema (Arllet)

(F) Chengzhi

(R) Roslen


15:00 		Coffee break

Entering data in ICIS Part 5: Ecology Management

ICISwbk-EMS: Extending ICIS to the farmer's reach (Warren)

Retrieving information from ICIS

The Zeus Web Application: Integrated Germplasm Browser and StudyMart (Mylah)

(F) Chengzhi

(R) Roslen

Day 4 -

Friday,
5 Mar 2010

8:30

Retrieving information from ICIS (cont'd)

Datawarehouse Tables (Ching)
ICISwbk-FBK: Harnessing the statistical power of R (Thomas)
Installing R and RExcel

ICIS housekeeping tools:

ICIS Administrator Tool (William)
Ontology Management: Uploading GRC and BPI characterization scores (Warren)


Monitoring ICIS developments:

Ensuring stability with Software Testing and capturing everything in and about ICIS through the ICIS Documentation System (Warren)

(F) Thomas

(R) Warren

10:30
 Coffee break

Release Management & Deployment of ICIS/MBP software

Overview of issues (Thomas)
Discussion

(F) Thomas

(R) Ching

12:30
 Lunch
13:30

Future Integration of ICIS and Genetic Resources Management

Overview of issues (Ruaraidh)
Discussion

Back to the Future: Quo Vadis ICIS?

Presentation (Richard)
Discussion
15:30
 Coffee break
16:00

ICIS Priority Setting for 2010 - 2011 (Thomas)

 


DISCUSSION

Day 1

Activity Reports of ICIS Partners

IRRI (Thomas)
CIMMYT (Arllet & Hector)
AAFC (Shawn c/o Arllet)


BAYER (Ed)
  • (presentation)
  • Q&A:
  • NO UPDATES BUT MORE OF A QUESTIONS
    • in bayer the research team pushed on intergrating molecular breeding platform and
      breding has been reorganized as center for global hands of breeding...
    • we need molecular marker breeding platform
    • formulating breeding nursery planner to manage their nurseries
    • what does a breeder do: selects a planting material, plants/regenerates,
    • needs more customizing
    • developed our own interface and application because we need a working interface
    • we became quiet, icis in crisis is relatively true for us in bayer
    • On in-house software : not sure if we can share it but the default is to keep it to ourselves
    • keep open the discussion lines between bayer and icis community
    • to make it work we would invest in our own internal effort
    • CHenzi : after organizing our new...i hope we can attract you back
    • Ed: im glad you said that
    • Graham : you said that some part of it is being developed in java right
    • Ed : i think it is
      • now May is indoctrinated with breeding needs/logic so somehow she manage to supply and give what breeder need
      • May and Casper are in the driving seat to cater breeders needs..


CBWA (Cameron)
  • (presentation)
  • Q&A:


New Members Key Interests in ICIS

UA (Howard Eagles)
GQNPC-IRRI (Roslen)
C4-IRRI (Richard)
  • The C4 rice project at IRRI is a Bill & Melinda Gates Foundation funded project to convert rice - a C4 photosynthetic crop - into a C4 photosynthetic crop. This is a complex biological engineering project which is still only in its basic early stage of plant genomics and physiology experimentation.
  • At IRRI, there are two components of the project that leverage ICIS
    • The IRRI team is generating, planting and characterizing plant physiological parameters of rice and sorghum mutant stocks, for the purpose of gene discovery. For this purpose, two ICIS Local databases have been set up (thank you Ching Habito!) with Zeus ICIS web interfaces (thank you Mylah Anacleto!).
      • This exercise turned out to be the first use of Zeus web with ICIS local databases. Zeus did not automatically map Central data into the data display. The CRIL team is fixing the GCP ICIS DataSource to fix this, but in the short term, Ching mirrored IRIS Central data (and simulated a Sorghum Central with rice data - but relevant sorghum data expected soon out of ICRISAT)
    • The C4 rice project will be generating and characterizing transgenic rice with C4 genes. The IRRI Rice Transformation Laboratory Information Management System (LIMS) is currently an Excel spreadsheet file based system, not tightly integrated with ICIS (even though the lab output is transgenic germplasm). This is not robust. A relational database tied to ICIS is a desired objective. Thus, CRIL has one software engineer working on the reverse engineering of the Excel LIMS into a relational database, the outputs of which will be cross-referenced in ICIS.
  • Q&A:
ICRISAT (Trushar c/o Arllet)

Q&A:

ISMAB

  • Howard Eagles: when can this be integrated with ICIS? this is what we want
  • on ISMAB- is there an assumption of genetic variance, are they qualitative traits or quantitative?
    • if you know all of them, you can get a prediction at the bottom
  • Graham: Guy is working on a tool called MOSEL
  • This is the first step, designed with a lot of discussion, with breeders in mind
    • density of markers
    • Ed: density of markers not ...
  • G: we have these extremes that we are dealing with: 1.I need to see all 5,000 SNPs 2.(other preference)

Installation of IChIS


GCP/MBP (Graham)
Thomas : i was under the impression that the icis/mbp would be entirely web-based?
Graham: two cases to support: web based, standalone
Richard: what is the "icis community" definition today? where does it need to go to become a robust professional community?
Graham: in mbp, we believe that the long-term sustainability depends on support and acceptance of the CG system.
Richard: there was a community exactly like the icis community. there were some dynamics there that i see in the icis community...how can we avoid that? let's look at other large open-source projects that actually survived...business model for prosperity. robustness of end-user installation..
Ruaraidh: is everything because of emails during the past few weeks?
Graham: no, because of events that transpired during the past year.
Ruaraidh: when you look at the definition of players, where do we (grc) fit in?
Graham: you would fit under the "CGIAR Lead Crop Centers".
Ruaraidh: however, we are not involved in the development of the configurable workflow, etc.
Graham: yes, the line needs to be drawn clearer.
Richard: how about the cgiar change process?
Graham: the cg has decided on 8 mega-programs (at concept level)..there is a meeting at the end of this month. we have open discussions with directors of centers.
Ed: we are not a small to medium enterprise. how do you organize mega-projects (like the ones big companies handle)? first rule of business is not to earn money but to sustain. You should tap resources from the private sector.
Graham: the steering committee has members from corporate organizations. corporate organizations have been successful in managing molecular breeding platforms. mission is, how can we emulate that and make it available for the public sector.



Day 2


Starting to use ICIS

Marketing ICIS through the new ICIS Customer Webpage (Warren)

ICIS Customer Page

  • on http://beta.irri.org/mainnew/products/icis-home/
  • Thomas: wrapping of live pages useful? squeezing the cropwiki pages awkward, there are a lot of information from cropwiki. Same case for linking to cropforge. It would be better to open a new browser tab to show the cropwiki and cropforge websites
    • Graham: why not put short instruction of how to submit bug report and then put link that would take you to the actual page
  • Thomas: use of logos: can we use the logo and just provide a link. Is this an endorsement of product
    • Thomas: persons present in this workshop to ask their legal department whether the use of logo is legitimate
    • this webpage is going through review and will be officially part of IRRI's web presence, this may be against the use of logos of partners
      • general problem in a network of community, you host in a server and becomes part of the institutional presence
    • Ed: okay as long as logo appears to be the same and not in any way modified.
  • Thomas: who would like to join the editorial, at least 4 should be on editorial board
    • CIMMYT: Guy, Hector, Juan Carlos
    • GCP-MBP: Graham
    • IRRI:
  • Feedback: the site is a good start
  • Graham: right from the beginning we chose icis.cgiar.org not irri.org



New IRIS Installer (Ching)
ICISwiki: NSIS Installer for IRIS 5.5
Comments and suggestions are welcome: http://www.cropinfo.org/icis/index.php/Talk:NSIS_Installer_for_IRIS_5.5
  • Q&A:
  • Process:
    • where to install it
      • Graham: have you tried testing on a drive other than the default
      • Ed: experience in the past: it is problematic, ensure successful installation on drives other than default
      • Richard: give user a synopsis of what other processes are (in the initial stage)
      • Graham: put text "this is step x"
  • Ed: installer questions, why do you need these kind of questions
  • Arllet: comments about asking user the level of access and who are the users
  • Ed: think like a novice user and not assume user knows
  • Thomas: will this installer lead to a production installer?
    • we are not only targeting users who do not know anything about ICIS but also the existing users
  • Graham: consult others, print each screen, ask feedback
    • screenshots with explanation/narrative about what that step is intended for
  • Chengzhi: go to the web and find other commercial systems, pattern after it, they usually have a better process, these products already went through the review process
  • Howard: how do we go around problems encountered in installation
  • Components to install
    • Graham: user may not know what to install here
    • Thomas: if it works for 40% of users, it should be sufficient?
    • Ed: 90%
      • simple and custom installation
    • Ed: shared drive location crucial
    • Ed: local databases not always on local machine but will be on shared drives. For us this is how it should be done
    • Thomas: the part when you would like to ask that is the initial screen, software installation target specified, where to put the data is another step
    • Graham: we walked right in the major complaint - whether to make it simple
    • Grace: it is not included in the installer, even after ODBC setup
      • Graham: should it be in the installer?
    • Ros: i was tempted to point the local db somewhere else. point central to another machine that has an updated central db
    • Ros: 2 persons came up with different data in the reports (in the SMTA column)
    • William: practice in IRRI of copying the central database; announces changes;
    • Ros: most updated central db - risk, might not be able to read announcement, forgot the copy
    • Thomas: opens up problem to be out of sync, not technical problem but management issue
  • Graham: does it install the training database as well - yes
    • by default installed
    • Ruariadh: option to not install training database? use of word "default" means there is another option
  • Graham: how easy is it for me to get that and install the package for chickpea
    • this crop may have a central, or may not
    • Arllet: how was it done for the case of installing IWIS .. that's what we need to do for chickpea
    • clear instructions to take this script.. etc
    • Arllet: installanywhere no need to recompile
      • for NSIS...
  • Graham: How is this related to Candy's installer
    • Arllet: installanywhere can help in installing mysql
    • Ching: I think this can include MySQL installation
    • Thomas: there is a section on release management, discuss later
  • Graham: part of the solution for this can be to ask questions first and execute the steps
  • Thomas: NSIS installer is open source and providing this kind of information allows to make custom installers
    • vs installanywhere that costs around $6000
    • to buy another license? NSIS a workaround to that bottleneck


GCPwizard + File Conversion Wizard: Migrating into ICIS the easy way (Warren)
  • Q&A:

First presn: migrating to ICIS the easy way

  • Arllet: can we change the keyword
    • Warren: currently hardcoded, not sure if we can add feature, project has ended
  • Graham: encountering a blank cell is dangerous
    • any way to indicate...
    • we do not want users leaving half of it unfilled
  • Graham: we already do it for INGER
    • general problem loading multiple dataset is workbook (general ICIS problem)
  • Richard: who uses it?
    • where has this tool been deployed so far?
    • (Rome collaborator) contacted principal investigators
    • Graham: very little feedback, got one email from Thailand
    • Thomas: Dr Seraj may be interested
    • Graham: what to do now with it
  • Richard: this was a GCP product
  • Graham: we are worried that we are making an excuse not an aid
    • encourage PIs to make data suitable and available for use
    • Thomas: a technical solution to a non-technical problem
    • Thomas: generally we should be cautious for providing such
    • Graham: wonderful product but how do we get people to use it
    • Ruariadh: too much and too little, this facility is providing too much but what are they getting out of it
    • Graham: can be intimidating
    • Ruariadh: if we can simplify it
    • Arllet: all treatment in one category rather than
  • Graham: before we put in more money we would like to see a serious user using it


Next presn

  • Ed: Would it allow you to
  • Arllet:
    • (1) don't give an option to convert to list, separate wizard for each
    • (2) when I make a mistake, going through the same step is tiresome, provide a back button
    • if you have a lot of files with the same structure, make use of a macro - that is the purpose of the record conversion button




Customizing ICIS

Development of Generic GRIMS (Juan Carlos)
  • Options: IMS, GRIMS, GRIN-GLOBAL, Development of a new system (in java platform)
    • Juan Carlos: We suggested to Genebank that GRIMS is probably the best option
    • Ed: better if there is a single global
    • Grace: we are not part of the project, will not be consulted
      • Ed: but why?
      • Ed: there is a lack of consortium, not a participatory mgt system
      • Ruariadh: exactly our problem
      • Ruariadh: they are tied in to milestones
      • Ruariadh: finish this project first then...
  • decision to use GRIM
    • Arllet: when dicussed with Guy, decided to try for 1 year, if not favorable use (GRIN-GLOBAL?)
    • Ruariadh: single seedbank management system
      • Ed: big players - chinese, japanese, what are they using
      • many systems but none of them are connected to breeding


Linking GRIMS and ICIS for Seed Multiplication (Beth)
  • Ed: last year you had the most structured process flow, i can relate to that
    • do you have a quality management system?
    • from there whatever it will be, we will be generating the forms
    • Ed: 5% lacking, we stepped back, will be putting quality mgt system in place (not ICIS)
    • Thomas: does it mean you think generic system does not work?
      • Ruariadh: we are very conservative in our ambitions
      • Graham: oracle use - not aware
      • Ruariadh: we cannot use MS Access, no expertise in ICIS for postgresql
      • Thomas: one of the defining criteria was you needed two systems running in parallel, can be done in oracle
      • it was not anticipated that you were going to use oracle
      • Ruariadh: this year switch off IRGCIS
  • Graham: i am interested that you decided to use central right away
    • Arllet: because for SMTA, we need data in central
  • Programmed in Delphi
  • Arllet: even if in oracle, it should have been easy to convert to postgresl
  • Beth: now converting to a more generic sql so conversion is easy


Separating Setgen into two use cases? (Ruaraidh)

problem is finding the source of that GID

  • Arllet: for incoming seeds, how about giving it another name type (e.g. rep seed health no.)
    • Graham: push it back giving some sort of stock no. from seed health
    • Arllet: through this, can directly reference to seed health

i do not see a single bloated SetGen to handle the two use cases

  • Graham: it seems to me that the major difference with it is in the first use case is importing germplasm and the other is progressing...
    • original plan was to have a pedigree input tool
    • eventually the progression tool can be streamlined to be easier to use
    • part of config workflow system
    • have not gone far in the external pedigree input tool
    • may be time to think about it again
    • Ruariadh
    • Howard: a common thing I do I put something it and take a lot of time determining the parent
      • not always have a lot of options
      • Ruariadh: yes, not special case with Genebank, use cases the same for others
    • Arllet: work on the same interface
    • Howard: just get rid of all othe other options
    • Arllet: Menu for creating cross, etc
    • Ed: exactly what we are doing in Bayer, all elements of SetGen are there but quality mgt system in place
    • Grace: a step by step procedure
      • Ed: should be up to the app mgr to customize for user
      • Grace: needed to study the setgen to be able to use it
      • we use the (-) button to advance 1 generation when in fact we are multiplying it
      • Graham: terminology issue
    • Arllet: I know Warren is transferring it to workbook
      • before we were able to make setgen, it took a long time to develop, test, etc
      • what took it run well is a lot of testing with user
      • Graham: nobody doubts that SetGen is not working well, so about ICIS Lite - a dazzling array of buttons.. discuss more later
Development of the Configurable Workflow (Chengzhi)
  • Arllet: "configurability"
    • many other terms e.g. multiplication, advancing the lines
    • Thomas: a generationalization - there are many ways of referring to the concept
    • Ed: big translation is in the...
    • Chengzhi: we can provide a configuration panel so the user can save that file to the user's file list, next time automatically load
  • Ed: feedback - when you want it to be successful to the intended user, back to the basics
    • e.g. what we need is a seed lot admin system
    • we failed to configure or make it work for us as a basic seed lot system
    • in India, much records, pure logistic report tool
    • it does not matter where they are, seed block propagation method
    • Chengzhi: user-driven, that is the approach
      • Chengzhi: but ideally we need the basic workflow
      • Ed: i am sure that for each breeding program basically the same
      • Ed: if it does not work for us it will not work for a lot of people




Supporting User Workflows through the ICIS Lite (Warren)
  • (presentation)
  • Q&A:
  • Graham: so a while ago we say that for icis 5.5 we have a a wide array of buttons
  • Warren : i set one value to off
  • Graham: you could give us
  •  Thomas: this is generally availablein an MS office feature. you have extended that a bit, turning off and turning on some icons
  • Graham : one issue is did we have this running in an indian computer?
  • Arllet : yes
  • Graham:what is ?yes we did or no we did?
  • Arllet: we run it  but it didn't work
  • Graham : what's the difference between 2003 qnad 2007 office?
    • so why does Hector have a workbook that says 2007?
  • Warren: what i did was to trap the error
  • Graham:so you do not have to maintian 2 versions?
  • Arllet: but why do i have a version that doesn't work?
    • Hector modified the old verison of the workbook and now it works
  • Warren: is the problem the menu?
  • Arllet: we don't know.errors occur in an uncertain manner
  • Graham: i was using 2007 and i have no problem
  • Thomas: there are different language versions of 2007 office, they have a french version and a spanish version
  • Arllet: how do you intend to make it work in all versions of office?
  • Warren: what we need to do is know what the problem is
  • Graham: since we are discussing it,will someone tell me if it is working in APU?
  • Arllet: Weng and i didn't manage to make it work..but later on Manoj said it was running well
    • but i didn't have the chance to ask Hector..he changed some thing then it works..
  • Warren: maybe he doesn't have the version that works
  • Thomas:are we planning to make it run in a mac?
  • Graham: well if we have someone come here and let it make it work in his mac then we have to
  • Thomas: i am asking this question becuase do we need to support these kinds of problems we had a problem of testing it in different environments because we don't have enoguh resources
  • Arllet: Richard was asking do we need to change OS to make it work?
  • Graham: but it is working in my office 2007 and windows 7
  • Arllet: that's the problem Graham, there has to be a combination of that makes this problem occur
    • this is always a problem that comes up whenever we are in the station..
    • i guess we will need to talk about how to do about it..
    • this problem will occur everytime..
  • Graham: probably this is not the time to discuss about this
  • Graham: do you experience a slowdown on office 2007 and 2003? because i experience that
  • Warren :haven't experienced it yet
  • Graham: now this is a challenge..copy the maize fieldbook menus
  • Warren: ok i have been doing that with the fieldbook
  • Graham: not everybody likes it but they know that
  • Arllet: i have problems with functions here..i am struggling what your functions meant
  • Graham: i think the function names are similar to the dll function names
    • i think we need to have a function name and what it does..i presume you have documented your code..
  • Arllet: the reason why i want it to be posted in cropforge is that we need to do some changes and we need to have an updated vesion in cropforge..
  • Thomas: so we can we discuss who should be able to customize this menu
  • Graham: the application managers
  • Arllet: i could be one, casper,...
  • Thomas: so those would be the experienced icis users
  • Arllet: how do you test the functions? because i had problems when testing functions that are dependent on another..so you need to test the functions even in a different order
  • Graham: we call it context sensitive bugs
  • Arllet: test it independently..independent of each other..it doesn't have to necessarily follow a pattern
  • Ros : are these bugs? because i intend to make some customizations myself so i need to know
  • Warren : i will discuss that later...

Entering data in ICIS Part 1: Genealogy Management

Managing documents and images (Grace)
Admin utility for handling persons, institutes, addresses, and locations data (Grace)
SetGen and GMSsrch UI changes (Ching)
GMS Schema 5.6 (Grace)


Day 3

Entering data in ICIS Part 1: Genealogy Management (cont'd)

ICISwbk-GMS: Managing germplasms with ease and comfort (Warren)
  • (ICISwkbk-GMS Tutorial Videos, by Aien) (poster)
  • Q&A:
  • Tool to generate cross
    • Arllet: Does the tool to generate cross check if there is already a cross created given the same female and male?
    • Warren:  Not sure but we will check the algorithm and include that.
    • Arllet: What crossing methods are supported by that method?
    • Warren:  The method in the training document is the one implemented.
    • Graham: Is that a full dialel?
    • Warren: Yes
    • Graham: Can you include more crossing options
  • Input Parser
    • If there are several records of the same name, it is suggested to have it with different color from those germplasm with one hit when searched from the database
  • Issues about problems encountered in different Office and Windows
    • Test the ICIS Lite in the version of Office and Windows that are currently used by the majority of the users, the latest released Office and Windows. At this point, it should Office 2003, 2007, 2010 in Windows XP, 2007 and Vista
    • Technical support during training through remotely accessing the user's computer with tools like WedEx, GoToMyPC etc
    • Install ICIS in the user's computer and ask them to  run some major tools of Workbook like retieval of the study
  • Who will test the application
    • Gouyou will be the main person to give feedback through his communication with breeders or ask a breeder to test it
  • Availability of the codes
    • Put the codes in the CVS
    • Strict versioning



Entering data in ICIS Part 2: Inventory Management


ICISwbk-IMS: Tracking everything through barcoding (Warren)
  • (presentation)(poster)
  • Q&A:
  • Beth : Have you considered the size of the labels?
  • Warren : It is the responsible of the users to define the labels
  • Arllet: I think it is good to have default lay-out and default sizes already available for user to choose
  • Juan Carlos: What's the relation between the old and new Intrack?
  • Warren : New Intrack uses the IMS database
  • Arllet : Where do you store viability data?
  • Warren : It's in the DMS
  • Arllet: Who will be the possible people to test it?
  • Warren: Bibiana, Group under Transgenic Lab, C4 group etc.
  • Arllet: Is there any documentation?
  • Warren:  It is the immediate thing we will do, to have the documentation available soon.

Entering data in ICIS Part 3: Data Management

ICISwbk-DMS (Warren)
  • (presentation)
  • Q&A:
  • Graham : It will be helpful if you could add more descriptions to the function windows that appears, so users will be guided what to choose.
  • William : Will it be possible to trap errors to avoid for the workbook to resset?
  • Warren : It should be but I still need to determine where to put in the trap.
  • Arllet: I think it is better to have a key that turn off the Debug error message and have a friendly message if error occurs if the Workbook is deployed in the user's computer. Then just turn it on when it needs to be debug.
  • Arllet: I noticed that the errors in Workbook usually occurs with the Cancel and close button.  Also, it happens when the state of your Workbook file is not compatible with what the particular tool needs to do. So, check the state of the Workbook before running the tool.
  • Arllet:  Is there a tool to generate bardcode after a fieldbook is done, like barcode for plot?
  • Graham:  The barcode is already generated for seed lot of the Inventory, so you can use the same approach. 
  • Warren:  That can easily be done.


ICIS Mobile: Taking ICIS out into the field (Warren)
  • (video, by Edward) (poster)
  • Q&A:
  • Addition of valid scale values
    • there should be a capability for the user to record values out of the valid values and record it as valid value for that session
  • Rauraidh: Barcoded values to be supported in reading by the mobile
  • Setting up of the factor and variate
    • Graham: Is there only one factor that will be loaded in the mobile sheet?
    • William : One is the minimum
    • Graham: Why is the factor set-up in the computer and the variate in the handheld?
    • Arllet: The user should not be asked to select database
  • Defining a row or observation unit.
    • Graham: If a unit needs to be defined by several factors, you should allow that flexibility.
    • Arllet:  Sometimes the breeders measure by block, so it should allow arranging it by block and so forth
  • Grace: Does it allow you to measure 2 or three variables?
  • William : Yes, by selecting the variate in the list box.
  • Graham and Arllet:  It should allow the breeder to measure a plot then go through three variables he choose. Then it will automotically move to the next plot.
  • Rauraidh: It should give a descriptive information about the plot like name of the germplasm.
  • Arllet: The order in the handheld should be the barcode or factor first, then the variable and the value.
  • Graham: The barcode should not be the label but someting plot number or entry number, depending on the name of the factor that was barcoded. 

Entering data in ICIS Part 4: Gene Management

ICISwbk-GEMS (Warren)
  • (presentation)
  • Q&A:


Notes:

  • There is an error filling up the marker name field.
    • Maybe use a marker ID?
    • There should a protocol to follow in adding a marker into the database.
    • Guy is proposing a new development for GeMS
      • Particularly for big studies
  • Other concerns
    • Parallel loader
      • works but is slower
      • speed is an issue in storing and retrieving data
    • how to store 2 alleles in 1 marker
      • separate them by comma in a cell
      • use a data matrix
    • DArT data is no longer possible to load using workbook -- it is just too large
    • Transformation is used to serialize data, but it is no longer possible to serialize when there are 2 alleles
  • Graham: ToDo's for GeMS:
    • Find a partner to work with in the use of the workbook
    • Arllet to provide the data from the ACIAR project (can use the data from the ACIAR project in India)
MBP - ICIS 6 Schema (Arllet)
  • (presentation)
  • Q&A:
  • Presentation URL: http://www.cropinfo.org/icis/index.php/TDM_Gene_Management_System_6.0
  • Domain model and schema on a 1-to-1 relationship
    • "entity" in the domain model is "gems-entity" table in the schema
  • Chengzhi: How to represent sequencing in the schema?
    • Richard: Answer is something to be explored by Ken and Chengzhi
  • Arllet: GeMS data are actually stored in a Study
  • Chengzhi: How to handle genotyping data without GID
    • Graham: It would be of little use to store genotyping data without GID
    • Graham: Accession numbers may be use to map the genotyping data without GID into some germplasm in ICIS
  • Ontology [1]
    • Arllet: There are issues with the ontology loader
    • Richard: What ID to use to link Ontology with ICIS?
      • Richard: The ontology ID may not necessarily be invariant
      • Arllet: Character ID is not good to use because it makes things run slower
      • Guy, Richard, Arllet, Chengzhi could work on finding a solution
  • Graham: ICIS 7 may not necessarily be backwards compatible with ICIS 5


Entering data in ICIS Part 5: Ecology Management

ICISwbk-EMS: Extending ICIS to the farmer's reach (Warren)
  • (presentation)
  • Q&A:
  • Graham: Is this module going to be useful for ICIS?
    • Chengzhi: Is it possible for them (the government agency consulting with CRIL) to do the programming?
    • Graham: It is a bit of a stretch to accommodate the pest data into ICIS
    • Paul Fox: Determine the priority before committing into developing this application
    • Graham: This is not a priority for the ICIS team right now
    • Thomas: This is somebody else's domain
    • Thomas: ICIS is an overkill for this application
    • Graham: We have a lot more to do in the molecular breeding platform than tow orry about this application
    • Paul Fox: The issue is more on the Rice Knowledge bank than as an issue for consideration in ICIS development
    • Warren: IRRI entom department is coordinating with me on this matter
  • Chengzhi: Are there action items for this?
    • Graham: Wait for future developments. It is not a priority for now.

Retrieving information from ICIS (cont'd)

The Zeus Web Application: Integrated Germplasm Browser and StudyMart (Mylah)


  • Graham: Paul Fox is interested to see the new version
  • Mylah: A bit of history of Zeus
    • First released in 2007 with Dr. Bruskiewich
    • Several seminars organized by Thomas to involve the users in planning out the application
      • geared in improving the interface
      • CIMMYT people gave their comments about the software
    • Live version is already deployed in a server at IRRI
  • Chengzhi: The web application searches across the entire DMS data
  • TODO:
    • Graham: Make the results in a tree structured form
    • Implement something for neighborhood search
    • Chengzhi: Searching is actually based on BioMart
    • Graham: Allow user to setup customized search filters
      • Thomas: Allow even searching for NULL valued fields
      • Graham: For numeric fields, implement searching between given minimum and maximum values- mylah: these are related to data curation issue
      • Thomas: The logical condition for filtering should be as configurable as possible
  • Graham: Provide textual guides in the dialog boxes so that users would know what to do every step of the way.
  • Paul: ICIS Web is putting us backwards
    • Paul: Howard Eagles was able to make scientific achievements using the standalone version of ICIS
    • Paul: Expose the new ICIS users to the stand-alone version and let the Web version develop, mature, and catch up with the stand-alone version

Day 4

Retrieving information from ICIS (cont'd)

Datawarehouse Tables (Ching)
  • (presentation) (poster)
  • More information: http://www.cropinfo.org/icis/index.php/ICIS_Data_Warehouse_Creation_and_Maintenance
  • Q&A:
    • Graham: What is the maximum number of datapoints being supported?
    • Thomas: Maybe hundreds
    • Ching: I'm not sure right now, I'll get back to you on that
    • Richard: Is there a way of incrementally increasing this tables?
    • Ching: Yes, we already started implementing a way to update each dataset only
    • Thomas: How long does it take to update the tables?
    • Ching: Just hours, not days
    • Arllet: Is there a way of checking the integrity of the data
    • Thomas: We encountered erroneous data in the core tables themselves, that should be cleaned first
    • Graham: Are the steps clear on how to handle the population of the tables?
    • Ching: I started putting them up in the Wiki
    • Thomas: Are the scripts under version control in CropForge?
    • Ching: Yes
    • Arllet: Why does the web need a warehouse in the germplasm details?
    • Richard: Because of the pedigree viewer, there's a performance issue in generating the trees
ICISwbk-FBK: Harnessing the statistical power of R (Thomas)
  • Installing R and RExcel (video, by Dec) (poster)
  • ICISwbk-FBK (presentation)
  • Q&A:
    • Graham: Great progress!
    • Arllet: Is it not hard to integrate R?
    • Thomas: No, you download the executable installer and just accept the default setup. After installing R, you download the needed R graphic libraries (i.e. iPlots) then download RExcel as well


ICIS housekeeping tools

ICIS Administrator Tool (William)
  • (video, by Edward) (poster)
  • Q&A:
    • Graham: Would it be helpful to record the date when the last data validation has been done on the database?
    • William: Yes, we can use the INSTLN table to record that
    • Arllet: Would it be useful to record the history and not just the last date?
    • Graham: Yes, that's a good thing
    • Thomas: I just want to make it clear that we have placed the scripts in XML files so that it can be modified
    • William: Where do we put the edited data when you are editing the central?
    • Warren: It's possible to put it in the central directly depending on the user access rights

Monitoring ICIS developments

Ensuring stability with Software Testing and capturing everything in and about ICIS through the ICIS Documentation System (Warren)
  • (presentation)
  • Q&A
  • Graham:  It is great to have those types of documents specified. However I doubt about the sustainability of it if those documents scattered in different places only accessible to certain people. Making it shareable and having other people involved in the document will make it more sustainable as to write those documents require great efforts. So we should look into the different tools available to manage these documents.  There is cropforge to manage codes. Should the description of the functions in the development manual in the code itself?  The whitebox document can be a wiki page which can be filled up as developer and bug reporter discuss further about it and then link it with the bug report. Based on Warren's presentation, probably what he needs is more powerful collaborative tools
  • Richard: There is JIRA now available for us to manage projects which Guy uses. Probably, it is high time for us to commit in using it and the entire development teams of CRIL should be using only one suite of tools. But we can further discuss this when we get to IRRI. Jay already set-up it up well and it is ready for immediate use.



Release Management & Deployment of ICIS/MBP software (Thomas)
  • (presentation)
  • Discussion:
    • Thomas: Its a nuisance getting the Workbook codes to/from Cropforge
    • Graham: when submitting a bug report, we can send a patch (of code) that will fix the bug if we are able to view the source code.
    • Thomas: (warren and arllet for example) everyday you would have to download the other's sourcecode and import that sourcecode before making your changes.communicate..as a workaround to importing and exporting all the time via CVS to keep in sync with what the other is doing
    • Graham: need to consider people who dont have commit privileges
    • Arllet: (re release naming/numbering) we also need to consider the schema(including warehouse)-datasource-webapp version compatibility for the ICIS web.
    • Chengzhi: how about we agree on the versioning number across the three teams?
    • Graham:release managers are:
      • 1) web --> Mylah/Chengzhi
      • 2) eclipse rcp --> configurable workflow: Guy
      • 3) traditional --> Warren
    • Graham: top two items (above) share a lot of backend technology, chengzhi responsible for backend for item 1 and 2
    • Thomas: we have 3 release managers...we might have applications where we will require little coordination (some teams may release more frequently than others).
    • Graham: matrix of cril releases...(see photo below)
T-B (left column): web, configurable workflow system (CWS), standalone (SA) ICIS
    • Graham: whenever you want to update your central...
    • Chengzhi: we should enforce updating the central and local at the same time.
    • Thomas: upgrading is the applications & schema,updating is the data
    • Graham: ching, (for NSIS) 1) I would like to have something that others can easily change for other crops/creating own installers
    • Graham: 2) i'd like to see a version that can handle mysql as well.
    • Chengzhi: too dangerous for me for the installer to install mysql as well. why not just install mysql manually?
    • Richard: we looking to moving away from ms access, and distributing data in mysql. we need to think this through carefully.
    • Thomas: have diagnostic run first (check if JRE, MYSQL is installed, etc.) before installing...have messages to suggest things that need to be installed or upgraded, etc.
    • Graham: there is an intermediate solution...for people with connectivity: when someone becomes a user of the platform, they get their own virtual server somewhere. next step down is to translate that to distributing that physically to someone (via CD/usb, etc.)
    • Thomas: previously the INstallshield installer could you everything. It's logical to separate these functions into an installer (for fresh installs), and upgrader. we want to reduce the options / questions being asked esp for new users (therefore install the training database automatically in a crop-specific installer).
    • Graham: administrator's task: packup new central, packup new local for a user, tell user to download this and run it.
    • Chengzhi: the installer should ask server locations, etc. Give them the requirements for ICIS to run before installation/upgrade/update.
    • Thomas: how do you confront users with questions like "would you like to replace the JRE with a new one?" consider implications that could possibility break his/her machine.

Future Work

Future Integration of ICIS and Genetic Resources Management (Ruaraidh)
  • Presentation (Ruaraidh)
  • Discussion
  • 2010 Changes
    • FileLink->Bibrefs  Reflink for documents
    • FileLink->Attribute for images (Note: Lose 2-level hierarchy group of images)
    • Separate EVENT table which we can call PROJECT
Back to the Future: Quo Vadis ICIS? (Richard)
  • Presentation (Richard)
  • Discussion
    • Graham: the tool most difficult to integrate to the eclipse/rcp is the Excel worksheet/fieldbook. the end-user still wants the Excel spreadsheet.
    • Thomas: Edward has explored java libraries that not only interact with excel but with Openoffice. Spreadsheets will just be containers for data, no macros.
    • Thomas: we have been able to integrate R into excel, but no progress yet in integrating R in the eclipse/rcp
    • Thomas: Richard, you cannot refer to the ICIS standalone tools as "legacy". these tools are being used in the MBP for gaining users.
    • Graham: Ideas were thrown around that problems will be solved by stopping the development of the standalone version.
    • Richard: Existing standalone tools are actually serving the community very well. Especially with the all the work that Warren's team has put in during the past year.
    • Graham: the things that have to catch up: 1) GEMSCAT. 2) Gems module of Workbook. If we are to have another GEMSCat, it should be programmed into the eclipse/rcp and using GEMS schema 6.0
    • Thomas: How about using the Record Keeper functionality of the ICIS Administrator tool (now being developed in eclipse/rcp) as vehicle for delivering the new GEMScat?
    • Graham/Chengzhi: (re moving ICIS into "the cloud") There are policies within our power to offer but not to enforce it.
    • Thomas: after working out our own issues on our own servers,vision is to streamline this on other servers we manage?(where we deliver our product)
    • Chengzhi: how about if the users insist on using their own servers?
    • Richard: thats ok as long as they can support themselves.
    • Graham: re Deployment problems: # of users using icis, also using analtytical tools (some directly within icis, some not direct)..we should find a central place where we would install software, and users may access via own accounts. local database would be here. they would access a shared Central database, etc.
    • Thomas: that's the nunhems solution.
    • Graham: yes, thats what i would like for the MBP. i dont know how many servers we need, how many installations of mysql we need, etc.
    • Richard: you buy what you need. depends on the model you pick...
    • Graham: i also have to think about people without connectivity. usb stick solution?
    • Thomas: which legal entity makes the contract with google to have this kind of service? who is paying the bill? who will share the cost?
    • Richard: that is the third part of my presentation...
    • Thomas: i think there is a third option, the "icis foundation". establishing a foundation: they have seen that as a way for software development, attracting funding.
    • Thomas: when you use the word "business" i'm thinking of profit-making.
    • Richard: I meant "business" in the sense that it's structured, not a loose collection of individuals. aim for a professional society, a legal entity. can employ people, etc.
    • Thomas: advantage of foundation would be abiding by a Constitution. there is the sense of the long-term view.
    • Richard: time has come to put a stronger structure in place.
    • Graham: we have a window of opportunity to put things in place.
    • Thomas: would that be done via the CGIAR?
    • Richard: again, have to consider the cgiar change process currently going on. Maybe now is the time to bring this up with CG management.


ICIS Priority Setting for 2010 - 2011 (Thomas)
Old Work from 2009

See 2009 Future work inventory.

New Work in 2010
  • ICIS Database Structure
    • (CRIL) – Changes table for DMS (Arllet, Ching, Warren)
    • (CRIL) – Identify maintenance methods in the Germplasm table (GNPGS=-2) (ICIS Developers)check applications test .i.e. GNPGS < 0 or = -1
    • (CRIL) – Create Central ICIS DB with PostgreSQL and use it for web interface.(Ching, Mylah, Richard)
    • (CRIL/GRC) - Location Module (Grace, Ching)
      • DLL catch up with location table changes (Warren)
      •   Ruaraidh: to complete managing mission information
    • (CIMMYT) Joint review of ICIS & GRIN Schemata to identify points of cross linkage (Grace, Bibiana, Juan Carlos)
    • Ontology curation terms (Graham, Rosemary)
    • Ontology schema integration with ICIS 5.6 (Arllet, Richard, Ching)
    • Ontology DLL (Warren, Graham)
    • GEMS (Guy, Trushar) 
    • High density genotyping data (Guy, Trushar)


  • MBP CWFS (Chengzhi, Mylah, Guy, Jeffrey)
    • Local and Central concept in GCP data source (Jeffrey Morales)
    • Implement write access to database (Jeffrey, Guy)
    • Implement basic authentication in GCP datasource(Jeffrey, Guy)
    • Implement connection to R modules (Guy)
    • Concrete applications:
      • StudyMart (Mylah, Chengzhi)
      • Maintenance of Germplasm and study browser (Mylah, Chengzhi)
      • Pedigree/neighborhood viewer/curation/editing tool (Martin, Kevin)


  • ICIS Database Access
    • (CRIL/AAFC) Multi-user Access and Restriction in Postgres (Shawn,Ching)
      • Security/access of data on ICIS level (Warren)
    • (GRC/CRIL) GRIMS Backend conversion of Oracle to PostgreSQL (Ruaraidh)


  • Stand-alone Applications
    • (CRIL) - cross finding facility in standalone applications (Ching, Graham)
    • (CRIL) - administrator’s tool (Edward, William)
      • Loading/Deleting Central Studies
      • Partial upload of Germplasm records
      • Synchronization of GMS and DMS GIDs
    • (CRIL/AAFC) - implementation of naming conventions (Ching, Shawn)
    • (CRIL) - Improvement of InTrack application (Juan Carlos)
    • (CRIL/GCP) - Development of MoSel (Guy, Teri)
    • (CRIL/CIMMYT) - finalizing version 9 of Fieldbook with links to ICIS (Vivek, Hector, Claudio)
    • (CRIL/INGER) - Seed ordering tool (Lance, Ed)
    • (GRC) Enhancement of GRIMS - Location, People, Institute
    • (CRIL/GRC) Incoming seeds tool (Chengzhi, Ruaraidh, Roslen)
    • (IRRI- GIS) GPG2 Location data curation and validation (Grace)
    • (CRIL) ICIS Lite (Warren, Dec, Aien)


  • Content Management/Documentation
    • (ICIS community) - ICIS website, ICISWiki, Confluence, JIIRA, LinkedIn (Vivay, Guy)
      • Rice (William)
      • Cookbook for wheat (Arllet -> NN, Shawn, Sandra)
      • Maize (Hector, Juan Carlos)
      • Breeding training/multimedia materials (Vivay)


  • ICIS implementations and conversion
    • (CIMMYT/CRIL)-look at the transfer of wheat genetic resources data to ICIS following with the experience on rice (Juan Carlos, Bibiana)
    • (IRRI/CIAT/WARDA/IITA/GRIN)-GPG2 cross referencing of rice germplasm, data loading (Grace)
    • (CIMMYT/ICARDA/GRIN)-GPG2 cross referencing of wheat germplasm, data loading  (Juan Carlos, Arllet, Weng)
    • (UQ/AAFC/ICARDA/CRIL) – Uploading pedigree updates into IWIS3 and IBIS (Shawn, Sandra, Arllet)


  • Release Management
    • InstallAnywhere - MySQL installation (Candy)
    • NSIS - customized installer scripts for use by others (Ching)
    • Source code management (team leaders, Chengzhi, Guy)
ICIS deployment
  • WARDA, Africa (Ibnou Dieng)




  • Review ICIS GMS schema tables relating to genetic resources (IRGCIS) conversion
    • Review 2009 data model discussion (and GCP domain model) - review in a CRIL teleconference (in near future?)
      • What are the "use cases" (SQL queries)
        • How many are GRC managerial versus end user queries?
      • Affected tables:
        • LISTNMS
        • EVENTMEM table
        • Collecting Missions(?)
        • FILELINK table - ICIS REFLINKS table(?)




PARTICIPANTS


Venue: University Club, University of Western Australia
No Name Affiliation
1 Thomas Metz CRIL (Crop Research Informatics Laboratory), IRRI
2 Warren Constantino CRIL, IRRI
3 Maria Corina Habito CRIL, IRRI
4 Nigel Ruaraidh Sackville Hamilton GRC, IRRI
5 Grace Lee Capilit GRC, IRRI
6 Lilibeth Sison GRC, IRRI
7 Arllet Portugal CRIL, CIMMYT
8 Juan Carlos Alarcon CRIL, CIMMYT
9 Graham McLaren GCP (Generation Challenge Programme)
10 Roslen Anacleto GQNPC (Grain Quality, Nutrition and Postharvest Center), IRRI
11 William Eusebio PBGB (Plant Breeding, Genetics and Biotechnology), IRRI
12 Chengzhi Liang CRIL, IRRI, CIMMYT
13 Mylah Rystie Anacleto CRIL, IRRI
14 Richard Bruskiewich CRIL and C4 Rice Photosynthesis Project, IRRI
15 Cameron Beeck University of Western Australia / Canola Breeders Western Australia
16 Adam Hunter Murdoch University
17 Paul Fox Australian Center for International Agricultural Research
18 Ed Roumen Bayer CropScience
19 Howard Eagles The University of Adelaide
20 Takiyaudin bin Haji Mohamad Department of Agriculture & Agrifood, Brunei Darussalam
21 Hirman bin Haji Abu Department of Agriculture & Agrifood, Brunei Darussalam








Retrieved from "http://www.cropinfo.org/icis/index.php/ICIS_Developers_Workshop_Australia_2010"
Personal tools