Difference between revisions of "WormBase-Caltech Weekly Calls"

From WormBaseWiki
Jump to navigationJump to search
 
(360 intermediate revisions by 11 users not shown)
Line 19: Line 19:
 
[[WormBase-Caltech_Weekly_Calls_2018|2018 Meetings]]
 
[[WormBase-Caltech_Weekly_Calls_2018|2018 Meetings]]
  
 +
[[WormBase-Caltech_Weekly_Calls_2019|2019 Meetings]]
  
GoToMeeting link: https://www.gotomeet.me/wormbase1
 
  
  
= 2019 Meetings =
 
  
[[WormBase-Caltech_Weekly_Calls_January_2019|January]]
+
= 2020 Meetings =
  
[[WormBase-Caltech_Weekly_Calls_February_2019|February]]
+
[[WormBase-Caltech_Weekly_Calls_January_2020|January]]
  
[[WormBase-Caltech_Weekly_Calls_March_2019|March]]
+
[[WormBase-Caltech_Weekly_Calls_February_2020|February]]
  
[[WormBase-Caltech_Weekly_Calls_April_2019|April]]
+
[[WormBase-Caltech_Weekly_Calls_March_2020|March]]
  
[[WormBase-Caltech_Weekly_Calls_May_2019|May]]
+
[[WormBase-Caltech_Weekly_Calls_April_2020|April]]
  
[[WormBase-Caltech_Weekly_Calls_June_2019|June]]
+
[[WormBase-Caltech_Weekly_Calls_May_2020|May]]
  
[[WormBase-Caltech_Weekly_Calls_July_2019|July]]
 
  
[[WormBase-Caltech_Weekly_Calls_August_2019|August]]
+
== June 4, 2020 ==
  
 +
=== Citace (tentative) upload ===
 +
* CIT curators upload to citace on Tuesday, July 7th, 10am Pacific
 +
* Citace upload to Hinxton on Friday, July 10th
  
== September 12, 2019 ==
+
=== Caltech reopening ===
 +
* Paul looking to get plan approved
 +
* People that want to come to campus need to watch training video
 +
* Masks available in Paul's lab
 +
* Can have maximum of 3 people in WormBase rooms at a time; probably best to only allow one person per WB room
 +
** Could possibly have 2 people in big room (Church 64) as long as they stay at least 10 feet apart
 +
* Need to coordinate, maybe make a Google calendar to do so (also Slack)
 +
* Before and after you go to campus, you need to take your temperature and assess your symptoms (if any) and submit info on form
 +
* Also, need to submit who you were in contact with for contact tracing
 +
* Form is used all week, and hold on to it until asked to be submitted
 +
* If someone goes in to the office, they could print several forms for people to pick up in WB offices
  
=== Update on SVM pipeline ===
+
=== Nameserver ===
* New SVM pipeline: more analysis and more parameter tuning
+
* Nameserver was down
* avoiding precision (and F-value) as a measure (dependent on ratio of positives and negatives in test set)
+
* CIT curators would still like to have a single form to interact with
* For example shown, "dumb" machine starts out with precision above 0.6
+
* Is it possible to create objects at Caltech and let a cronjob assign IDs via the nameserver? May not be a good idea
* G-value (Michael's invention); does not depend on distribution of sets
+
* Still putting genotype and all info for a strain in the reason/why field in the nameserver
* Applied to various data types
+
* We plan to eventually connect strains to genotypes, but need model changes and curation effort to sort out
* Analysis: 10-fold cross validation
+
* Hinxton is pulling in CGC strains, how often?
** Randomly select 10% pos and neg (without replacement) and repeat until all papers sampled
+
* Caltech could possibly get a block of IDs
* F-value changes over different p/n values; G-value does not (essentially flat)
 
* Area Under the Curve (AUC): probability that a random positive scores higher than random negative
 
* AUC values for many WB data types upper 80%'s into 90%'s
 
* Ranjana: How many papers for a good training set? Michael: we don't know yet
 
* Can't reproduce old training sets (for old SVM); provide Michael better training sets if you want improved SVM
 
* If SVM still not good enough, Michael will work on deep neural networks (Tensor Flow)
 
* Michael can provide training sets he has used recently
 
  
=== Clarifying definitions of "defective" and "deficient" for phenotypes ===
+
=== Alliance SimpleMine ===
* WB phenotype ontology has many "variant/abnormal" terms and distinct subclass terms for "defective/deficient"
+
* Any updates? 3.1 feature freeze is tomorrow
* Have tried to create a logical definition pattern for these terms, but the vagueness of the meaning of "defective" and how it is distinct from "abnormal" has stalled the process
+
* Pending on PI decision; Paul S. will bring it up tomorrow on the Alliance PI call
* What do we mean exactly by "defective" and how, specifically, is this distinct from "abnormal"?
+
 
* Definitions include meanings or words:
+
 
** "Variations in the ability"
+
== June 11, 2020 ==
** "aberrant"
+
 
** "defect"
+
=== Name Service ===
** "defective"
+
* Testing site now up; linked to Mangolassi
** "defects"
+
* CGI from Juancarlos not accepting all characters, including double quotes like "
** "deficiency"
+
* Example submission that fails via CGI
** "deficient"
+
WBPaper000XXXX; genotype: blah::' " ` / < > [ ] { } ? , . ( ) * ^ & % $ # @ ! \ | &alpha; &beta; Ω ≈ µ ≤ ≥ ÷ æ … ˚ ∆ ∂ ß œ ∑ † ¥ ¨ ü i î ø π “  ‘ « • – ≠ Å ´ ∏ » ± — ‚ °
** "disrupted"
+
* Juancarlos will look into and try to fix
** "impaired"
+
 
** "incompetent"
+
=== Alliance Literature group ===
** "ineffective"
+
* Textpresso vs. OntoMate vs. PubMed
** "perturbation that disrupts"
+
* Still some confusion about what the different tasks can be performed in each tool
** Failure to execute the characteristic response = abnormal?
+
* Working on collecting different use cases on spreadsheet
** abnormal
+
* Sentence-based search is big strength of Textpresso
** abnormality leading to specific outcomes
+
* At latest meeting performed some large searches for OntoMate and Textpresso
** fail to exhibit the same taxis behavior = abnormal?
+
* Literature acquisition: still needs work
** failure
+
** Using SVM vs. Textpresso search to find relevant papers
** failure OR delayed
+
** Species based SVM? Currently use string matching to derive different corpora
** failure, slower OR late
+
** Finding genes and determining which species those genes belong to?
** failure/abnormal
+
 
** reduced
+
=== Alliance priorities? ===
** slower
+
* Transcription regulatory networks
 +
* Interactions can focus on network viewer eventually
 +
** May want different versions/flavors of interaction viewers
 +
** May also want to work closely with GO and GO-CAMs
 +
* Gene descriptions can focus on information poor genes, protein domains, etc.
 +
 
 +
=== Sandbox visual cues ===
 +
* Juancarlos and Daniela will discuss ways to provide visual cues that a curator is on a sandbox form (on Mangolassi) vs live form (on Tazendra)
 +
* AFP and Micropub dev sites have indicators
 +
* Could play with changing the background color? Maybe too hard to look at?
 +
* Change the color of the title of the form, e.g. the OA?
 +
* Will add red text "Development Site" at top of the OA form
 +
 
 +
=== Evidence Code Ontology ===
 +
* Kimberly and Juancarlos have worked on a parser
 +
* Will load into ACEDB soon
 +
 
 +
 
 +
== June 18, 2020 ==
 +
 
 +
=== Undergrad phenotype submissions ===
 +
* Chris gave presentation to Lina Dahlberg's class about community phenotype curation
 +
* Class took survey about experience with presentation and experience trying to curate worm phenotypes
 +
** Survey results: https://www.dropbox.com/s/00cit5aitv8yu27/Dahlberg_class_survey_results.xlsx?dl=0
 +
** Some students didn't benefit, but most did; nice feedback!
 +
** Lina intends to publish/micropublish the survey results so please don't share
 +
* Since April 24, the class has submitted 171 annotations from 23 papers (some redundant and some still under review)
 +
 
 +
=== Special characters in OA/Postgres ===
 +
* There are many special characters in free text entries in the OA; probably all from copy-pasting directly from PDF
 +
* In some cases it seems the special characters cause problems for downstream scripts (e.g. FTP interactions file generator)
 +
* It would probably be good to script the replacement of special characters with their appropriate simple characters or encoded characters
 +
* Juancarlos wrote Perl script on Mangolassi at:
 +
** /home/postgres/work/pgpopulation/grg_generegulation/20200618_summary_characters/get_summary_characters.pl
 +
** Will find bad characters and their pgids for a given Postgres table
 +
** Will find bad data and their pgids for the same table
 +
** People can query their data tables for these characters
 +
* Chris & Wen will work on compiling a list of bad characters that tend to come up
  
 
=== Citace upload ===
 
=== Citace upload ===
** Tuesday, Sep 24th
+
* July 10th citace-to-Hinxton upload
 +
* July 7th citace upload, but Wen will be on vacation so will upload to Wen on Tuesday, June 30th
  
=== Strain to ID mapping ===
 
* Waiting on Hinxton to send strain ID mapping file?
 
* Hopefully we can all get that well before the upload deadline
 
* Will do global replacement at time of citace upload (at least for now)
 
  
=== New name server ===
+
== June 25, 2020 ==
* When will this officially go live?
+
 
* Will we now be able to request strain IDs through the server? Yes
+
=== Caltech Summer Student ===
 +
* Paul has new summer student
 +
** Molecular lesion curation, maybe
 +
** Are early stops more or less likely to be null mutations?
 +
** Alleles are flagged as null in WB in the context of phenotypes
 +
** Would be good to query Postgres for null alleles and work from there
 +
* Fernando
 +
** Anatomy function
 +
** GO curation? Curating transcription factors?
 +
*** Checking for consistent curation
 +
 
 +
=== Worm Community Diversity Meeting ===
 +
* Organized by Ahna Skop and Dana Miller
 +
* Invite posted on Facebook "C. elegans Researchers" group
 +
* Two meetings held: one Thursday (June 18th), one Friday (June 19th)
 +
* Chris attended last Friday (June 19th)
 +
* Worm Board looking to take input and ideas from this meeting and incorporate into meetings and events
 +
* One idea was to document and track outreach efforts and what people have learned from them and organize them in a central location, maybe WormBase or Worm Community Forum
 +
* Also, there was a suggestion to have a tool that could inform potential students of worm labs in their respective local area
 +
** Ask Todd; he used to have a map of researchers; Todd had asked Cecilia to curate lab location and institution
 +
* Person and Laboratory addresses in ACEDB have a different format, so looking to reconcile
 +
* Do we know how many labs are still viable? Check for a paper verified in the last 5 years, or requested strains from the CGC recently
 +
** Most labs were real
 +
 
 +
=== C_elegans Slack group ===
 +
* Called "C_elegans"
 +
* Chris made a "WormBase" channel for people to post questions, comments
 +
* Chris will look into inviting everyone and possibly integrating with help@wormbase.org email list
 +
 
 +
=== WormBase Outreach Webinars ===
 +
* While travel is still restricted, we should consider WormBase webinars
 +
* Scott working on a JBrowse webinar
 +
* Could have a different topic each month
 +
* Should collect topics to cover and assign speakers (maybe multiple speakers per topic; keep it lively)
 +
* Should set up a schedule
 +
* How should we advertise? Can post on blog, twitter, etc.
 +
 
 +
=== New transcripts expanding gene range ===
 +
* Will bring up at next week's site-wide call
 +
* Possibly due to incorporation of newer nanopore reads
 +
* Many genes coming in WS277 have expanded well beyond the gene limits as seen in WS276
 +
** Example genes: pes-2.2, pck-2, herc-1, atic-1
 +
* Has several repercussions:
 +
** WormBase does not submit alleles affecting more than one gene; with these gene expansions suddenly alleles once only affecting a single gene are now affecting two genes, and so are now omitted from loading into the Alliance (including any phenotype and/or disease annotations)
 +
** Some expanded genes are now being attributed with thousands of alleles/variants
 +
 
 +
=== Citace upload ===
 +
* Upload files to Spica/Wen by Tuesday (June 30th) 10am
 +
* Wen will clean up folders in Spica (older files from WS277 not cleared out for some reason)
  
=== SObA Graphs ===
+
==July 9th, 2020==
* New graphs now live on site (Expression, Gene Ontology, Human Diseases, Phenotypes)
+
===Gene names issue in SimpleMine and other mining tools===
* A lot of whitespace padding above and below graph; maybe trim? trimming vertically would ultimately limit the view pane when user wants to zoom in, so we should leave as is for now
+
*Wen: Last week, Jonathan Ewbank raised the issue of gene names that may refer to multiple objects.
* Diff tool: Raymond and Juancarlos created a prototype diff tool (for comparing two genes, for example)
+
*this can be an issue for multiple data mining tools including WormMine, BioMart, and Gene Set Enrichment.
** Paul: compared two genes that should be very similar, but there are a lot of differences; may reflect annotation coverage rather than biology
+
*Perhaps have a standalone approach to check if any gene name among a list may refer to multiple objects (users check their name lists before submitting them to any data mining tool).
 +
*Jae: The public name issue has heterogeneous natures. That means there may be no single solution to solve all those problems.
 +
*Gene list curation from high-throughput studies, confusing usage of public names probably less than 2% (still cannot be ignored). See examples below--
 +
**single public name is assigned to multiple WBgene ID, Wen has a list of these genes
 +
**overlapped or dicistronic genes, ex. mrpl-44 and F02A9.10
 +
**overlapped or dicistronic, but has a single sequence name, examples:
 +
    exos-4.1 and tin-9.2 (B0564.1)
 +
    eat-18 and lev-10 (Y105E8A.7)
 +
    cha-1 and unc-17 (ZC416.8)
  
 +
**simple confusion from authors, ex. mdh-1 and mdh-2
 +
*One of the most significant problems is a propagation to other DB and papers of  these gene name issues.
 +
*We can make a special note for each gene page, but the people using batch analysis could not catch that easily.
 +
*Conclusion: Jae and Wen will work on a tool that lets Users "sanitize" their gene lists before submission to data mining tools.  They will also write a microPub explaining this issue to the community.
  
== September 19, 2019 ==
+
===Wormicloud===
 +
*Please test and leave any feedback on the word cloud tool (Wormicloud), https://wormicloud.textpressolab.com/
 +
*Valerio and Jae have worked on a tool that uses data in Textpresso; given a keyword, eg. "transposon", the tool generates a word cloud and word trend.
 +
*Any keyword can generate a graph that plots trends of occurence across the years in publication abstracts.
  
=== Strains ===
+
===Noctua 2.0 form ready to use===
* Need to wait for new strain IDs from Hinxton before running dumping scripts
+
*Caltech summer student will try using Noctua initially for dauer (neuronal signaling) pathways
* Don't edit multi-ontology strain fields in OA for now!
 
* Juancarlos will map free text strain entries to strain IDs once we have the complete mapping file
 
* "Requested strain" field in Disease OA; not dumped, so don't need to worry about right now
 
  
=== Alliance literature curation ===
+
===Nightly names service updates to postgres===
* Working group will be formed soon
+
*Nightly using Matt's wb-names-export.jar to get full output of genes from datomic/names service, and updating postgres based on that.
* Will work out general common pipelines for literature curation
 

Latest revision as of 22:11, 9 July 2020

Previous Years

2009 Meetings

2011 Meetings

2012 Meetings

2013 Meetings

2014 Meetings

2015 Meetings

2016 Meetings

2017 Meetings

2018 Meetings

2019 Meetings



2020 Meetings

January

February

March

April

May


June 4, 2020

Citace (tentative) upload

  • CIT curators upload to citace on Tuesday, July 7th, 10am Pacific
  • Citace upload to Hinxton on Friday, July 10th

Caltech reopening

  • Paul looking to get plan approved
  • People that want to come to campus need to watch training video
  • Masks available in Paul's lab
  • Can have maximum of 3 people in WormBase rooms at a time; probably best to only allow one person per WB room
    • Could possibly have 2 people in big room (Church 64) as long as they stay at least 10 feet apart
  • Need to coordinate, maybe make a Google calendar to do so (also Slack)
  • Before and after you go to campus, you need to take your temperature and assess your symptoms (if any) and submit info on form
  • Also, need to submit who you were in contact with for contact tracing
  • Form is used all week, and hold on to it until asked to be submitted
  • If someone goes in to the office, they could print several forms for people to pick up in WB offices

Nameserver

  • Nameserver was down
  • CIT curators would still like to have a single form to interact with
  • Is it possible to create objects at Caltech and let a cronjob assign IDs via the nameserver? May not be a good idea
  • Still putting genotype and all info for a strain in the reason/why field in the nameserver
  • We plan to eventually connect strains to genotypes, but need model changes and curation effort to sort out
  • Hinxton is pulling in CGC strains, how often?
  • Caltech could possibly get a block of IDs

Alliance SimpleMine

  • Any updates? 3.1 feature freeze is tomorrow
  • Pending on PI decision; Paul S. will bring it up tomorrow on the Alliance PI call


June 11, 2020

Name Service

  • Testing site now up; linked to Mangolassi
  • CGI from Juancarlos not accepting all characters, including double quotes like "
  • Example submission that fails via CGI
WBPaper000XXXX; genotype: blah::' " ` / < > [ ] { } ? , . ( ) * ^ & % $ # @ ! \ | α β Ω ≈ µ ≤ ≥ ÷ æ … ˚ ∆ ∂ ß œ ∑ † ¥ ¨ ü i î ø π “   ‘ « • – ≠ Å ´ ∏ » ± — ‚ °
  • Juancarlos will look into and try to fix

Alliance Literature group

  • Textpresso vs. OntoMate vs. PubMed
  • Still some confusion about what the different tasks can be performed in each tool
  • Working on collecting different use cases on spreadsheet
  • Sentence-based search is big strength of Textpresso
  • At latest meeting performed some large searches for OntoMate and Textpresso
  • Literature acquisition: still needs work
    • Using SVM vs. Textpresso search to find relevant papers
    • Species based SVM? Currently use string matching to derive different corpora
    • Finding genes and determining which species those genes belong to?

Alliance priorities?

  • Transcription regulatory networks
  • Interactions can focus on network viewer eventually
    • May want different versions/flavors of interaction viewers
    • May also want to work closely with GO and GO-CAMs
  • Gene descriptions can focus on information poor genes, protein domains, etc.

Sandbox visual cues

  • Juancarlos and Daniela will discuss ways to provide visual cues that a curator is on a sandbox form (on Mangolassi) vs live form (on Tazendra)
  • AFP and Micropub dev sites have indicators
  • Could play with changing the background color? Maybe too hard to look at?
  • Change the color of the title of the form, e.g. the OA?
  • Will add red text "Development Site" at top of the OA form

Evidence Code Ontology

  • Kimberly and Juancarlos have worked on a parser
  • Will load into ACEDB soon


June 18, 2020

Undergrad phenotype submissions

  • Chris gave presentation to Lina Dahlberg's class about community phenotype curation
  • Class took survey about experience with presentation and experience trying to curate worm phenotypes
  • Since April 24, the class has submitted 171 annotations from 23 papers (some redundant and some still under review)

Special characters in OA/Postgres

  • There are many special characters in free text entries in the OA; probably all from copy-pasting directly from PDF
  • In some cases it seems the special characters cause problems for downstream scripts (e.g. FTP interactions file generator)
  • It would probably be good to script the replacement of special characters with their appropriate simple characters or encoded characters
  • Juancarlos wrote Perl script on Mangolassi at:
    • /home/postgres/work/pgpopulation/grg_generegulation/20200618_summary_characters/get_summary_characters.pl
    • Will find bad characters and their pgids for a given Postgres table
    • Will find bad data and their pgids for the same table
    • People can query their data tables for these characters
  • Chris & Wen will work on compiling a list of bad characters that tend to come up

Citace upload

  • July 10th citace-to-Hinxton upload
  • July 7th citace upload, but Wen will be on vacation so will upload to Wen on Tuesday, June 30th


June 25, 2020

Caltech Summer Student

  • Paul has new summer student
    • Molecular lesion curation, maybe
    • Are early stops more or less likely to be null mutations?
    • Alleles are flagged as null in WB in the context of phenotypes
    • Would be good to query Postgres for null alleles and work from there
  • Fernando
    • Anatomy function
    • GO curation? Curating transcription factors?
      • Checking for consistent curation

Worm Community Diversity Meeting

  • Organized by Ahna Skop and Dana Miller
  • Invite posted on Facebook "C. elegans Researchers" group
  • Two meetings held: one Thursday (June 18th), one Friday (June 19th)
  • Chris attended last Friday (June 19th)
  • Worm Board looking to take input and ideas from this meeting and incorporate into meetings and events
  • One idea was to document and track outreach efforts and what people have learned from them and organize them in a central location, maybe WormBase or Worm Community Forum
  • Also, there was a suggestion to have a tool that could inform potential students of worm labs in their respective local area
    • Ask Todd; he used to have a map of researchers; Todd had asked Cecilia to curate lab location and institution
  • Person and Laboratory addresses in ACEDB have a different format, so looking to reconcile
  • Do we know how many labs are still viable? Check for a paper verified in the last 5 years, or requested strains from the CGC recently
    • Most labs were real

C_elegans Slack group

  • Called "C_elegans"
  • Chris made a "WormBase" channel for people to post questions, comments
  • Chris will look into inviting everyone and possibly integrating with help@wormbase.org email list

WormBase Outreach Webinars

  • While travel is still restricted, we should consider WormBase webinars
  • Scott working on a JBrowse webinar
  • Could have a different topic each month
  • Should collect topics to cover and assign speakers (maybe multiple speakers per topic; keep it lively)
  • Should set up a schedule
  • How should we advertise? Can post on blog, twitter, etc.

New transcripts expanding gene range

  • Will bring up at next week's site-wide call
  • Possibly due to incorporation of newer nanopore reads
  • Many genes coming in WS277 have expanded well beyond the gene limits as seen in WS276
    • Example genes: pes-2.2, pck-2, herc-1, atic-1
  • Has several repercussions:
    • WormBase does not submit alleles affecting more than one gene; with these gene expansions suddenly alleles once only affecting a single gene are now affecting two genes, and so are now omitted from loading into the Alliance (including any phenotype and/or disease annotations)
    • Some expanded genes are now being attributed with thousands of alleles/variants

Citace upload

  • Upload files to Spica/Wen by Tuesday (June 30th) 10am
  • Wen will clean up folders in Spica (older files from WS277 not cleared out for some reason)

July 9th, 2020

Gene names issue in SimpleMine and other mining tools

  • Wen: Last week, Jonathan Ewbank raised the issue of gene names that may refer to multiple objects.
  • this can be an issue for multiple data mining tools including WormMine, BioMart, and Gene Set Enrichment.
  • Perhaps have a standalone approach to check if any gene name among a list may refer to multiple objects (users check their name lists before submitting them to any data mining tool).
  • Jae: The public name issue has heterogeneous natures. That means there may be no single solution to solve all those problems.
  • Gene list curation from high-throughput studies, confusing usage of public names probably less than 2% (still cannot be ignored). See examples below--
    • single public name is assigned to multiple WBgene ID, Wen has a list of these genes
    • overlapped or dicistronic genes, ex. mrpl-44 and F02A9.10
    • overlapped or dicistronic, but has a single sequence name, examples:
   exos-4.1 and tin-9.2 (B0564.1)
   eat-18 and lev-10 (Y105E8A.7)
   cha-1 and unc-17 (ZC416.8)
    • simple confusion from authors, ex. mdh-1 and mdh-2
  • One of the most significant problems is a propagation to other DB and papers of these gene name issues.
  • We can make a special note for each gene page, but the people using batch analysis could not catch that easily.
  • Conclusion: Jae and Wen will work on a tool that lets Users "sanitize" their gene lists before submission to data mining tools. They will also write a microPub explaining this issue to the community.

Wormicloud

  • Please test and leave any feedback on the word cloud tool (Wormicloud), https://wormicloud.textpressolab.com/
  • Valerio and Jae have worked on a tool that uses data in Textpresso; given a keyword, eg. "transposon", the tool generates a word cloud and word trend.
  • Any keyword can generate a graph that plots trends of occurence across the years in publication abstracts.

Noctua 2.0 form ready to use

  • Caltech summer student will try using Noctua initially for dauer (neuronal signaling) pathways

Nightly names service updates to postgres

  • Nightly using Matt's wb-names-export.jar to get full output of genes from datomic/names service, and updating postgres based on that.