present
Introduction by Rob Elshire
- Future problems solvers - introduce ideas without prejudice
- egalitarian meeting
- Setup a space on GitHub and Gitter
MBIE project is missing a bit that is the toolkit to capture the audit trail to record what is done.
Sorting this out... today is a technical discussion
Last year we spent time discussing this at MapNet 2014.
- Getting on GitHub was a barrier - action this before 11am
- Some time out/moved from AgResearch
- GitHub space needs organisation
- scripts from Cornell
- documentation from Mingsu
- pre TASSEL QC
- Mingsu has done some work
- Patrick Biggs (not present) has generated some
- Justin Borwitz @ANU also done some
- Need to integrate this work
- What itches need scratching
- personal interests
- job overlap
Setting up best practice and technology transfer.
Rob
- automatability
- reproducibility
- getting it out there and available
- CRI and NZGL platform
Rudiger
- work within the MBIE project
- end to end trace
Shane
- problem solver and programmer
Tony
- BioIT provision
- provide and environmnet that is useful
- organisation
- project management
Naumann
- What can Bioinformatics@AgResearch do as a team
- Team leader
Marcus
- what is going on? - save effort
- reproducible research
- unit testing and tool evalutation
- best practice
Mingshu
- wish unified platform for researchers and analysts
- small country - maximise leverage
- discoverability
Milan
- feed data and have a publication
- User and exploit
Roger
- data platform
- genome assembly
- understand science and plant breeding programmes and applications
Helge
- automation to make us competitive
- best practice
- Kea and integration for samp
Ruy
- new comer
- steps through pipeline
- QC
- multivariate stats approaches and effects
Rachel
- pipeline runner
- accuracy - are they working as expected
Aurelie
- Bacteria - new area
- learning new things
- using Github for reproducible science
Cecilia
- standardisation
- lab practice and effects of results
- best practice wet and dry labs
** Phil **
- delinearise
- VISG 2 informed decisions for project plan
- modularisation of a pipeline
- exome capture using same/similar tools
- assay tools for end user needs
- validation component
- best practice
- biologists require a level of confidence form output
** Rob **
- linkages and feedback with down stream analysis
- down stream and upstream communications
- solve the blindsiding
Samples are different...
- hackathon
- uniq pipeline is deprecated
- slow compared to
- small overlap
- small overlap between tools
- TASSEL workflow discussion - compressing reads -> tags -> reads per sample to genotyping
- sequencing depth per sample?
- not in scope
- discussion
- GBS X?
Rob drawing boxes on the whiteboard
- purple to match Phil's shirt ;)
- areas to work on
- capturing what is important
- QC
- demultiplexing
- SNP calling
- Testing
- Analysis
Across board is automation, reproducibility, best practice and documentation.
LIMS - > raw data -> VCF - > 'real' results
-
rabbit hole discussion about demultiplexing and snp calling.
-
1 - 3 have common ground as do the across the board - these are the core
-
documentation - need for wet and dry versions
-
moa == automation and reproducibility
Analysis
Phil asked direction of TASSEL 5 dev.
- TASSEL 5 use case is land races - out breeding
- Are they amendable to community derived code? - yes - a reason for BBS approach.
To enable some distributed workflows within git.
We are dealing with biology, different organisms have different volumes of resources, and hence best practice.
We could do with one.
What should they do.
Steering not oversight
- IP
- Contributor agreement
- Resourcing
Naumann and Helge are CRI go-to people.