WormBase-Caltech Weekly Calls November 2015

November 5, 2015

IEA GO Annotations informating descriptions for non-elegans

• Some non-elegans genes have confident-sounding descriptions, about localization for example; likely not manual
• Will make sure automated descriptions handle IEA's appropriately

Wen using Textpresso machine to build SPELL

• Athena was taking 48 hours to build SPELL (4GB max memory allotted)
• Textpresso machine took 24 hours (8GB memory allotted)
• Maybe we can use Textpresso machine for future builds
• SPELL java program running on Athena (alloted 4GB out of 6GB)
• Can monitor Athena memory usage; check for any memory leaks

Community Curation Tracker CGI

• Purpose: To automate many steps of the e-mailing process to authors requesting community curation and to track those papers for which we've sent e-mails
• There is a "Ready" form and a "Tracker" form; "Ready" for generating list of papers to e-mail about, "Tracker" to track the status of papers that have already been e-mailed about
• Wiki documentation at: http://wiki.wormbase.org/index.php/Contacting_the_Community#Community_Contact_.26_Curation_Tracker_CGI
• Sandbox CGI located at: http://mangolassi.caltech.edu/~postgres/cgi-bin/community_curation_tracker.cgi
• EMAIL FUNCTIONALITY IS LIVE: please note!
• Allele-phentoype("New mutant" datatype) CGI is live on Tazendra
• Concise description version will be cloned from the New Mutant CGI and modified as needed

SObA Graphs for phenotype

• Juancarlos will transfer code tomorrow, if there are no change requests before then
• Existing feature requests will be added to future releases of SObA

Micropublications

• Daniela spoke with Daniel about having web pages in WormBook style to house micropublications

Status of WormBase User Survey?

• Paul sent around paper this morning: http://www.nature.com/nature/journal/v527/n7576_supp/full/527S16a.html
• What aspects of databases really get used by users?
• How can we be more efficient with existing curation pipelines?
• Question of the month on WormBase?
• We could get user feedback when visiting labs

Sending author updates

• Would be nice to have a mechanism to update authors on what has been curated for their paper
• Could give paper data summaries as well as recent updates to curation for a paper

Objectifying/Itemizing annotations from Postgres

• What if we objectified itemized Postgres entries for easier display of all annotations per paper, for example?
• May require addition of a text field tag to the ?Paper class, for example, that would capture each annotation item for a paper
• This would be in lieu of actually objectifying all annotations

November 12, 2015

Concise description form

• URL for listing genes mentioned in a paper
• Ranjana will send to Kimberly

Community curation CGI

• Almost ready for prime time
• ~155 e-mails sent out for allele-phenotype; ~25-30 responses

Scripts for community description form

• James wrote scripts that will need to inform community CGI
• Ranjana will work with Juancarlos

FlyBase text mining

• Text mining for Fly corpus, entity recognition and extraction, esp. for transgenes
• Need to improve SVM on transgene and allele recognition
• Textpresso working on tools for FlyBase

Phenotype SObA graph

• Making progress towards making the graph a widget on the gene page
• Still issues to iron out; working with Sibyl
• Some conflicts with Cytoscape for interactions vs. SObA

Data Visualization program opportunity

• Will have to write a grant
• Would pay for one person, get two with grant
• Anyone with ideas should consider a proposal
• Dec 15th deadline

WormBook

• Daniel prepared mockup for new web page

Biocuration meeting

• Mary Ann working out plans for community curation workshop
• Does PomBase track their community curation?

HT115 vs HT115(DE3) E.coli strains

• Redundant, should probably merge HT115 into HT115(DE3), as the RNAi-host E.coli strain

Visualization plans

• Karen e-mailed around older proposal for work with Santiago Lombeyda
• Original project was to upgrade visualization of pathways and virtual worm in WB page

November 19, 2015

Community curation

• Community e-mailing for allele-phenotype: 315 e-mails sent, 49 papers curated, 260 community annotations in the Phenotype OA
• Chris gave Paul list of all sy* alleles and phenotypes
• Paul going through table to add phenotypes not already in WB
• Paul using Textpresso and an allele-phenotype worksheet
• Do we want to create more elaborate forms or parallel forms? Probably parallel forms (in future)
• How can we assess completeness of curation? Very difficult to automate, currently
  • Check figures and tables
  • Ask authors about completeness
• Publication templates/worksheets? Streamline data submission concurrent with publication
• Can discuss at SAB in Feb
• We can send PI's/authors lists of alleles and what phenotypes WB has

Working with NCBI

• Interested in Textpresso text-mining and SVM categorizing
• How might we work with NCBI & PubMed Central to curate? They are willing to run SVM on PMC corpus

Minimum information requirements

• RNAi MI seemed bulky and not complete
• Asks authors for genes targeted, not dsRNA sequence
• We could write up a paragraph outlining the complaints about the proposed RNAi MI
• Maybe micropublish the paragraph

Micropublications

• Should go into PubMed Central

Allele descriptions

• Karen wants to adapt Ranjana and James' automated gene descriptions for alleles
• Start working on in January

SObA

• Juancarlos making progress on integrating into the WB site
• There are difficulties getting SObA into it's own widget on the gene page
  • Conflicts with other settings (javascript / cytoscape) on same page, e.g. Interactions widget (w/Cytoscape viewer)
  • Updated cytoscape.min.js version on Interactions viewer

Corresponding author curation

• Juancarlos and Cecilia have worked out a pipeline to capture corresponding authors for each paper
• Once paper editor changes are decided, can go live

WormBook chapters for WormBase

• Each chapter written for users, but we likely want to avoid too much technical jargon
• We do want transparency
• We can separate the chapters into less-technical and more-technical, for easier reading