WormBase-Caltech Weekly Calls August 2017
From WormBaseWikiJump to navigationJump to search
- 1 August 3, 2017
- 2 August 10, 2017
- 3 August 17, 2017
- 4 August 31, 2017
August 3, 2017
- Progress (maintenance stuff) can be written as of last grant cycle
- Current projects and report the steady state
- Some authors not reporting CRISPR alleles with standard nomenclature
- Some in Paul's lab creating same knockin with same name (should be distinct). Mary Ann to update nomenclature guidelines.
- Some (e.g. Bruce Bowerman) knock-in GFP in frame and use RNAi against GFP to knockdown endogenous gene
- RNAi mapping pipeline not setup to handle this
- Could maybe use endogenous sequence near insertion site (not ideal, for off-target effects)
- Maybe best, once new ?Phenotype_experiment model is in place, to just refer to perturbed gene
- Has been less stable lately; Raymond has needed to restart machine
- Being served from Dell server (in Braun building)
August 10, 2017
- Neuronal wiring with Scott Emmons? Maybe not for this round
- Transcription regulatory networks? Is a priority
- Paul will assemble existing parts and send around for review
- InSilico entity linking project requires curator link checking - for missing links as well as for faulty links - bad links are already an issue when ambiguities exist and this will grow as an issue when we start marking up multiple mod entities in the same paper
- There is a brief summary table outlining the link URLs and predicted quality - this is a useful feature of the current system set up by Arun, we are creating the same type of feature in the new inteface
- Could Cecilia check links?
- New interface will come out in a few months; if anyone starts now, they'll have to relearn when the new interface comes out
- Eventually markup should be job of authors during the process of writing. Need an authoring tool - this can be made - there needs to be more discussion between wormbase, and insilico if this is going to be part of the insilico phase II or not
- Should we have a standard human readable names for sequence features?
- Should we consolidate several features associated with the same gene when displaying in network views?
- Is there an automated way to give them public names?
- Maybe name automatically, but then allow manual overwriting
- Use gene name when experimentally demonstrated connection
- Needs to be automated, rule-based
- Could name based on neighboring gene(s)
Person lineage graph
- Progress made in development, direct lineage running off of Datomic. All of it will need to be completed before it can go into production.
Docker container system
- Potentially very useful for WB web services
Consider content for NAR paper
- Send ideas to Raymond
- Input for the WB grant?
- Could discuss microreviews
- Automated descriptions good for genes with little info
- Human-written descriptions good for genes with lots of info; don't change too frequently
- Some have suggested including protein domain info
August 17, 2017
- Citace upload August 29th (most likely)
AGR 1.0 Disease data
- Discussion of cutting back scope to only gene-disease associations versus pushing back deadline
- Problems stem from the fact that gene objects have already been instantiated as objects with human-readable names, synonyms, etc.
- Objects like strains, genotypes, fish, etc. still need to be instantiated with symbols, synonyms etc. and so will be difficult to get set up for human-readable display and linking to MOD pages
- We've already pushed back the deadline twice before, so we probably don't want to do that again
- We can put up disclaimers or notifications that we are only showing gene-disease associations for now, but other associations (to genotypes, strains, etc.) will come soon
NAR paper community curation?
- Not much has changed about community curation fundamentally since last NAR paper
- Chris will write up update with specific numbers (response rates, annotation numbers, etc.) for NAR paper
Community curation submission requests
- Chris & Juancarlos updated request pipeline
- Now focusing just on first authors; allows more emails to go out each month
- Have sent 333 emails since beginning of August, have received annotations for 20 papers so far (6% response, but still early; average tends to be 16%)
- Sending out more emails should result in more responses in absolute number
Author First Pass (AFP) pipeline
- Karen brought up that the wrong author was emailed for AFP (Author First Pass)
- Probably an issue that last author happened to not be the corresponding author, may have thrown off automated detection of corresponding author
- To what extent do we and community benefit from the AFP pipeline?
- There may be cases where AFP flags a paper for a data type that was missed/false negative by SVM
- We can optimize/update the AFP form, leave it as is, or possibly abandon it
- Maybe Chris can take over over-seeing emails sent to authors for various requests
- Would like more specifics on future plans
- Chris will send Paul some details on future plans for community annotation
August 31, 2017
- Mitani has submitted some for allele-phenotype and sequence change; we will not store the raw data for the sequence reads
- we will keep for now, but re-evaluate more thoroughly in the coming weeks, particularly wrt possible links to the micropublication platform
- Paul will send around a version soon and will contact specific people if he needs more info on a certain section
topic lists for SPELL and SpellMine
- we should take advantage of GO ontology structure for browsing topic lists
- Jae will start at WB on the 1st of September
- he'll work on PPI and possibly phenotype curation first, will add GO curation a bit later