Difference between revisions of "WormBase-Caltech Weekly Calls"

From WormBaseWiki
Jump to navigationJump to search
m
m
Line 16: Line 16:
 
[[WormBase-Caltech_Weekly_Calls_July_2011|July]]
 
[[WormBase-Caltech_Weekly_Calls_July_2011|July]]
  
 
+
[[WormBase-Caltech_Weekly_Calls_August_2011|August]]
 
 
== August 4, 2011 ==
 
 
 
Leak in office
 
*Emergency showers don't necessarily have drains
 
*We need to make sure that things in offices are prepared for potential water leaks
 
*Production server still running
 
*Mangolassi down for the time being; will be back up when dry
 
*Should call physical plant 24-hour emergency line (ext. 4717)
 
*Feel free to check offices/labs above ours to assess risk of leak
 
 
 
 
 
Testing concise description OA on sandbox
 
*Community annotation form needs to be very simple (e.g. gene name and empty box only)
 
 
 
 
 
Transgenomics pictures
 
*Daniela spoke with Mihail about getting pictures from their database
 
*We should know exactly what the clone is (sequence)
 
*What are they using for anatomy annotation? they use standard Anatomy Ontology (Daniela double checked with Mihail)
 
*Linking to Yuji Kohara's database (NEXTDB)?
 
 
 
 
 
Gene file from Hinxton to populate Gene Tables
 
*Kimberly working with Mary Ann Tuli and Michael Paulini to make sure each gene names are cleaned up
 
 
 
 
 
Raymond
 
*Trying to install Amigo (for Ontology browsing)
 
*Working on SPELL
 
 
 
 
 
SVM
 
*Karen using author first pass list for priority
 
*2 out of 3 papers have not been flagged for phenotype; Why?
 
*Michael setup a way to tag papers on Textpresso Dev if they've been processed by SVM
 
*How do we make sure all papers go through SVM pipeline?
 
 
 
 
 
Replacing power supply on conference phone
 
*Probably cannot fix
 
*Need to buy new one
 
 
 
 
 
== August 11, 2011 ==
 
 
 
Physical Interaction model
 
*Separating out from general "Interaction" model
 
*Combining the Yeast-Hybrid classes with the other Physical interactions
 
*Setting up a merged model that captures all information from previous Y2H, Y1H, and other physical interaction objects
 
 
 
 
 
Genetic Interaction Ontology
 
*Working with Rose, Andrew and Julie from BioGRID
 
*Established a good working (first draft) ontology for genetic interactions
 
*Will make adjustments to current document based on neutrality functions
 
 
 
 
 
FlyBase GSA Markup
 
*Working well with FlyBase
 
*Incorporating new corrections to scripts and entity lists
 
 
 
 
 
 
 
== August 18, 2011 ==
 
 
 
 
 
Next Thursday (Aug 25) is the next upload
 
 
 
 
 
Clarifying Phenotype-to-Gene Associations
 
*Currently, the phenotypes displayed on a given gene page may or may not be attributable to the gene on whose page the phenotype is being displayed
 
*Proposal is to separate out experiments in which a single gene has been perturbed from those experiments in which 2 or more genes have been perturbed, so as to clarify which gene or genes a certain phenotype can be attributed to
 
*All the information needed to implement this clarification is already available
 
**Information can be collected and displayed on the fly (but this will slow down the website considerably)
 
**Information can be collected offline (on a per release basis?) and displayed accordingly without on-the-fly calculations
 
**A curation tag could be added to the data model (RNAi experiment) to capture whether or not the experiment is "Simple" (one gene is pertubred) or "Complex" (two or more genes are perturbed)
 
 
 
 
 
ACEDB vs Web Display
 
*To what extent should we consider modification of the database to make amenable for web display?
 
*Front-end vs back-end calculations and manipulations
 
*Needs to be worked out for the next grant cycle
 
 
 
 
 
GSA Markup Pipeline
 
*Do others want to volunteer to contribute?
 
*More heads could help make the GSA markup pipeline more efficient and faster
 
 
 
 
 
== August 25, 2011 ==
 
 
 
 
 
Citace upload today
 
 
 
 
 
Physical Interaction model
 
*Kimberly will send to the Models list
 
*Requesting comments on the model
 
*Feedback also on the idea of separating Interactions into Physical, Genetic, Regulatory, and Predicted Interaction models
 
 
 
 
 
NAR 2011 Paper working draft
 
*Karen sent out first draft
 
*Many updates from the progress report have been slashed for priority/space
 
*Focused on new data types/models
 
*Awaiting feedback and contributions
 
*3 page (published) limit
 
*Word/character limit?
 
*Human disease relevance can be trimmed down (Ranjana)
 
 
 
 
 
Expression clusters
 
*Wen has been working on data and model
 
*In ~2 weeks will be finished
 
 
 
 
 
Wen starts RNAi curation around mid-September
 
 
 
 
 
WormBase Beta site launches next week (Sept 1st)
 
 
 
 
 
Routing issue with WormBook
 
*Not much we can do; just send Sanger mirror site link out as needed
 
  
  

Revision as of 17:08, 1 September 2011

2009 Meetings


2011 Meetings

February

March

April

May

June

July

August


September 1, 2011

SPELL

  • Raymond and Wen worked on putting on Amazon cloud
  • SPELL running on Amazon cloud
  • What next?
  • Need to be careful about running a cloud instance
  • Elastic IP? No, not yet
  • What are the requirements for SPELL? Free instance is currently sufficient
  • Need more memory for the start-up process
  • Haven't yet figured out a way to do stress tests
  • SPELL does not have many users (~500 queries per month)
  • Free tier available for only 6 months and then gone
  • Need to determine what the build requirements are (needed for every build/release)
  • Only want one copy of it (so we aren't paying twice)
  • Single canonical AMI
  • First step, firgure out build requirements (need more memory during build?) and runtime requirements
  • Free instance 2GB memory (upgraded) for build, 600MB memory OK for runtime
  • What will it take to maintain one of these instances for a year?
  • Double up SPELL queries with the main WormBase instance on the cloud to consolidate?
  • How to more fully incorporate SPELL into WormBase?
  • Integrate SPELL into the new WormBase App? (branding, tracking access)
  • Running under Apache; Ruby-on-Rails
  • Could maintain cloud instance for now and see how it will function (and what it will cost) over the year
  • Better if elastic IP assigned (25 cents/day? Or free once attached to an instance?)
  • Can it run in a micro-instance? Yes
  • $175/year for cpu plus disk space and transfer
  • Switch credit cards (from Raymond's to Caltech)
  • Follow logs to assess likely costs later on