Synopsis

This page summarizes the IRRI-CRIL MBP Consultation with IRRI Breeders about the MBP on Monday, February 9th, 2009.

Agenda

Generation CP Gates proposal submission for a next generation Molecular Breeding Platform
Start an enhanced process of stakeholder-driven software system development
Just at the starting line…

Firefox versus IE
Browsing Trait names
Get germplasm by country - timeout
Get germplasm by name e.g.'pokkali', IRGC 15238 won't show the germplasm details (error)
Pedigree
- pedigree window: appears empty (usability)
- when user selects different germplasm, open another pedigree window showing the pedigree tree of the currently selected germplasm
- request to filter by germplasm method
- option to supress the the display of germplasm identifiers in the node of pedigree

feature request: back button takes user to previous screen/view
feature request: pedigree- option to open all / several generations
button to generate pedigree - move to top of germplasm details page
options for pedigree tree generation
- add option for orientation
- show method checkbox: clarify, consider changing label to 'germplasm method',
- consider putting in instructions
name searches for germplasm
- option to search from preferred name or other type of germplasm name
- search from all names
apply the ICIS name standardization (e.g. IR64, IR 64 should combine search results)
search by trait (consider synonyms)
- e.g flowering and 'heading date'- ICIS should know about the synonyms
results list - sort all data from all pages
study view: sort the values of the factors and variates
new search string mode for germplasm: 'search by cross' where parents of the cross is delimited by '/' --> get IR???
- this is available in ICIS Browse.fcr command
make timeout errors more user-friendly
tabs: selected tab and sub-tabs in different color than the rest
for hyperlinks: change cursor or color
phenotype catalog: incorrect values loaded for plant / response mimic
mutant search: show processing icon

Would like to retrieve all germplasm specific data across all studies for a given trait (perhaps, directly from a trait search result)
- Filter for specific range of dates
- Filter for specific locations

Given a specific list of germplasm, want to retrieve all the available data for all available studies (across several seasons) for a specific set of measured traits, to allow global comparisons
- Values of the same Factor across all available studies for a given (set of) Germplasm
- Having some kind of statistic measures & visualization of data distributions of trait values across all data retrieved
- You might wish to constrain queries across a limited set of studies meeting a given set of criteria, e.g. common location, the trait measurement method used, common developmental stage.
  - Mainly coarse filtering of data sets to retrieve enriched to the point where manual analysis (e.g. in an Excel spreadsheet) becomes tractable, perhaps for R statistical analysis (following standard algorithms, expressed as R modules?)

Regarding naming designations, it would be helpful to have a glossary to tell users what the designation refers to, i.e. what Institute or breeding program or equivalent

Question: why is the process of ICIS experimental planning and data capture for/from plant breeding so error prone?
- Dave Mackill: Grace (now William, Connie) have to spend so much time rechecking and correcting data in new crosses (might be "pre-ICIS" data errors, now controlled better with SetGen + workbook?) Pedigrees not entered by hand anymore... might not be a source of error anymore(?).

Need an internal marker polymorphism database (SSR - past; SNP - future)
- Gel pictures with all experimental meta-data (gel concentration, standards used, etc.) indexed by germplasm (e.g. IR 64) and markers (e.g. RM20), where the lanes are documented for their contents (G x M)
  - Markers by accessions - see Alleles
    - e.g. IR 64 has been run by many groups internally
- SNP's will be much easier to score
Although SNP's are coming, SSR's will still be used for some time to come
- SSR's more useful when calibrated against known, well characterized standards like IR 64
  - Would it be good to have at least a well characterized database of "standards"
  - What markers work best for genotyping, for given polymorphism

Institutional versus technical issue?
- Not much incentive to properly document experiments for secondary use? e.g. clearly labeled gel pictures

Integrate with current tools for querying OryzaSNP
Need an internal "local" OryzaSNP for ongoing genotyping of many additional lines genotyped

Visualize SNP alleles, for a limited set (e.g. 384) of SNP loci, in different colours, for distinct SNP alleles, visualized on a chromosome spread for two contrasting varieties, side by side

Tool to select an optimal subset of SNPs available from a larger standard set, for a small subset of parental germplasm (could be more than 2 lines)
- Given target donors for specfied traits, want all polymorphic markers
- SNP quality scores
- Evenly spaced markers across genomic chromosomal (LD) "bins"

Graphical genotyping tool to visualize current graphical (chromosomal) genotypes/phenotypes to decide at each generation of a pedigree, which germplasm do you select for advancing, that minimize the target introgression regions (foreground), while minimizing genetic background

Overview of suggested process:
- Brainstorming session on needs (whiteboard, simple enumeration)
- Prioritize for most critical needs
- Develop at least one (templated) use case on one function/need (as an example)