Difference between revisions of "WormBase-Caltech Weekly Calls"

From WormBaseWiki
Jump to navigationJump to search
 
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
+
[[WormBase-Caltech_Weekly_Calls_2020|2020 Meetings]]
  
 +
[[WormBase-Caltech_Weekly_Calls_2021|2021 Meetings]]
  
−
= 2019 Meetings =
+
[[WormBase-Caltech_Weekly_Calls_2022|2022 Meetings]]
  
−
[[WormBase-Caltech_Weekly_Calls_January_2019|January]]
+
[[WormBase-Caltech_Weekly_Calls_2023|2023 Meetings]]
  
−
[[WormBase-Caltech_Weekly_Calls_February_2019|February]]
+
==April 11th, 2024==
 +
*Caltech WS293 ace files ready for the upload
  
−
[[WormBase-Caltech_Weekly_Calls_March_2019|March]]
+
==April 4th, 2024==
 +
* Continued discussion on sustainability
 +
* CZI, single cell RNAseq for Alliance -> anything happening will be few months down the road
 +
** Data is still going to SPELL and enrichment analysis
 +
** Peter Roy asking about expression profile of a condition and find similar expression profiles (SPELL like analysis) but SPELL cannot currently deal with scRNAseq data. Wen says it is possible (regarding each cell group as an experiment). Can try loading the into SPELL. Doe it improve the function of SPELL? Only 5-10 datasets. These data are a bit different from bulk RNAseq.
 +
* Textpresso: good to have a presentation for other MODs to show Textpresso capabilities? Yes. Maybe during sprint review
 +
* Michael's presentation on LLMs - Named Entity Recognition (NER)
  
−
[[WormBase-Caltech_Weekly_Calls_April_2019|April]]
+
==March 14, 2024==
  
−
[[WormBase-Caltech_Weekly_Calls_May_2019|May]]
+
=== TAGC debrief ===
  
−
[[WormBase-Caltech_Weekly_Calls_June_2019|June]]
+
==February 22, 2024==
  
−
[[WormBase-Caltech_Weekly_Calls_July_2019|July]]
+
===NER with LLMs===
  
−
[[WormBase-Caltech_Weekly_Calls_August_2019|August]]
+
* Wrote scripts and configured an LLM for Named Entity Recognition. Trained an LLM on gene names and diseases. Works well so far (F1 ~ 98%, Accuracy ~ 99.9%)
  
 +
* Is this similar to the FlyBase system? Recording of presentation  https://drive.google.com/drive/folders/1S4kZidL7gvBH6SjF4IQujyReVVRf2cOK
  
−
== September 12, 2019 ==
+
* Textpresso server is kaput. Services need to be transferred onto Alliance servers.
  
−
=== Update on SVM pipeline ===
+
* There are features on Textpresso, such as link to PDF, that are desirable to curators but should be blocked from public access.
−
* New SVM pipeline: more analysis and more parameter tuning
 
−
* avoiding precision (and F-value) as a measure (dependent on ratio of positives and negatives in test set)
 
−
* For example shown, "dumb" machine starts out with precision above 0.6
 
−
* G-value (Michael's invention); does not depend on distribution of sets
 
−
* Applied to various data types
 
−
* Analysis: 10-fold cross validation
 
−
** Randomly select 10% pos and neg (without replacement) and repeat until all papers sampled
 
−
* 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 curation status form development needs use cases. ref https://wiki.wormbase.org/index.php/WormBase-Caltech_Weekly_Calls#February_15.2C_2024
−
* WB phenotype ontology has many "variant/abnormal" terms and distinct subclass terms for "defective/deficient"
 
−
* 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
 
−
* What do we mean exactly by "defective" and how, specifically, is this distinct from "abnormal"?
 
−
* Definitions include meanings or words:
 
−
** "Variations in the ability"
 
−
** "aberrant"
 
−
** "defect"
 
−
** "defective"
 
−
** "defects"
 
−
** "deficiency"
 
−
** "deficient"
 
−
** "disrupted"
 
−
** "impaired"
 
−
** "incompetent"
 
−
** "ineffective"
 
−
** "perturbation that disrupts"
 
−
** Failure to execute the characteristic response = abnormal?
 
−
** abnormal
 
−
** abnormality leading to specific outcomes
 
−
** fail to exhibit the same taxis behavior = abnormal?
 
−
** failure
 
−
** failure OR delayed
 
−
** failure, slower OR late
 
−
** failure/abnormal
 
−
** reduced
 
−
** slower
 
  
−
=== Citace upload ===
 
−
** Tuesday, Sep 24th
 
  
−
=== 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 ===
+
==February 15, 2024==
−
* When will this officially go live?
 
−
* Will we now be able to request strain IDs through the server? Yes
 
  
−
=== SObA Graphs ===
+
=== Literature Migration to the Alliance ABC ===
−
* New graphs now live on site (Expression, Gene Ontology, Human Diseases, Phenotypes)
+
==== Use Cases for Searches and Validation in the ABC (or, what are your common actions in the curation status form)? ====
−
* 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
+
===== Find papers with a high confidence NN classification for a given topic that have also been flagged positive by an author in a community curation pipeline and that haven’t been curated yet for that topic =====
−
* Diff tool: Raymond and Juancarlos created a prototype diff tool (for comparing two genes, for example)
+
*Facet for topic
−
** Paul: compared two genes that should be very similar, but there are a lot of differences; may reflect annotation coverage rather than biology
+
*Facet for automatic assertion
 +
**neural network method
 +
*Facet for confidence level
 +
**High
 +
*Facet for manual assertion
 +
**author assertion
 +
***ACKnowledge method
 +
**professional biocurator assertion
 +
***curation tools method - NULL
  
 +
===== Manually validate paper - topic flags without curating =====
 +
*Facet for topic
 +
*Facet for manual assertion
 +
**professional biocurator assertion
 +
***ABC - no data
  
−
== September 19, 2019 ==
+
===== View all topic and entity flags for a given paper and validate, if needed =====
 +
* Search ABC with paper identifier
 +
* Migrate to Topic and Entity Editor
 +
* View all associated data
 +
* Manually validate flags, if needed
  
−
=== Strains ===
+
=== PDF Storage ===
−
* Need to wait for new strain IDs from Hinxton before running dumping scripts
+
* At the Alliance PDFs will be stored in Amazon s3
−
* Don't edit multi-ontology strain fields in OA for now!
+
* We are not planning to formally store back-up copies elsewhere
−
* Juancarlos will map free text and ontology-name strain entries to strain IDs once we have the complete mapping file
+
* Is this okay with everyone?
−
* "Requested strain" field in Disease OA; not dumped, so don't need to worry about right now
 
  
−
=== Alliance literature curation ===
+
==February 8, 2024==
−
* Working group will be formed soon
+
* TAGC
−
* Will work out general common pipelines for literature curation
+
** Prominent announcement on the Alliance home page?
  
−
=== SObA Graph relations ===
+
* Fixed login on dockerized system (dev). Can everybody test their forms?
−
* Currently only integrating over "is a", "part of" and "regulates"
 
−
* Maybe we could provide users an option to specify which relations to include, or maybe just exclude "regulates"
 
  
−
=== Author First Pass ===
+
==February 1, 2024==
−
* Putting together paper for AFP
+
* Paul will ask Natalia to take care of pending reimbursements
−
* Reviewing all user input for paper
+
* Dockerized system slow pages (OA and FPKMMine). Will monitor these pages in the future. Will look for timeouts in the nginx logs.
−
* Asking individual curators to check input
 
  
 +
==January 25, 2024==
  
−
== September 26, 2019 ==
+
=== Curator Info on Curation Forms ===
 +
* Saving curator info using cookies in dockerized forms. Can we deploy to prod?
  
−
=== Strains ===
+
=== ACKnowledge Author Request - WBPaper00066091 ===
−
* Paul D generated WBStrains for the missing TransgeneOme objects
+
* I am more than willing to assist; however, the task exceeds the capabilities of the normal flagging process.
−
* Working on a pipeline to identify new TransgeneOme strains at each upload
 
−
* One TransgeneOme object had 2 strains. Possible solutions: dump 2 expression objects that differ only in the Strain or remove the UNIQUE tag in the data model
 
  
−
=== Community phenotype requests August 2019 ===
+
* The paper conducts an analysis of natural variations within 48 wild isolates. To enhance the reliability of the variant set, I utilized the latest variant calling methods along with a custom filtering approach. The resulting dataset comprises 1,957,683 unique variants identified using Clair3. Additionally, Sniffles2 was used to identify indels of >30 bp, which numbered in the thousands to tens of thousands for most wild isolates. It is worth noting that variants identified with Sniffles2 have less reliable nucleotide positions in the genome.
−
* Sent out new round of phenotype requests on August 20, 21, and 22, 2019
+
 
−
* 2,626 emails/papers requested
+
* I am reaching out to inquire whether WormBase would be interested in incorporating this dataset. An argument in favor is the higher quality of my data. However, I am mindful of the potential substantial effort involved for WormBase, and it is unclear whether this aligns with your priorities.
−
* 114 emails bounced; 5 resent to new addresses
+
 
−
* 460 Phenotype OA community annotations; 181 RNAi OA annotations (641 annotations total)
+
* Should WormBase decide to use my variant data set, I am more than willing to offer my assistance.
−
* From 94 papers (83 for Phenotype OA; 33 for RNAi; 22 for both)
+
 
−
* By 81 distinct community curators (70 for Phenotype OA; 32 for RNAi OA; 21 for both)
+
=== Update on NN Classification via the Alliance ===
−
* 50 papers flagged as not having phenotypes (40 papers DO have phenotypes; 10 marked as negative; 80% failure rate!)
+
* Use of primary/not primary/not designated flag to filter papers
−
* 4 papers flagged for phenotypes (only 2 had curatable phenotypes; 1 had honey-induced phenotypes)
+
* Secondary filter on papers with at least C. elegans as species
−
* 115 papers with responses (5% response); 24 papers with input that were not main focus of request
+
* Finalize sources (i.e. evidence) for entity and topic tags on papers
 +
* Next NN clasification scheduled for ~March
 +
 
 +
* We decided to process all papers (even non-elegans species) and have filters on species after processing.
 +
* NNC html pages will show NNC values together with species.
 +
* Show all C. elegans papers first and other species in a separate bin.
 +
 
 +
=== Travel Reimbursements ===
 +
* Still waiting on October travel reimbursement (Kimberly)
 +
* Still waiting on September and October travel reimbursements (Wen)
 +
 
 +
=== UniProt ===
 +
* Jae found some genes without uniProt IDs, but the genes are there on uniProt but without WBGene IDs.
 +
* Wen reached to Stavros and Chris to investigate WormBase and AGR angles.
 +
* Stavros escalates the issue on Hinxton Standup.
 +
* Mark checks Build scripts and WS291 results. After that, he contacted UniProt and he's working with them to figure this out.
 +
 
 +
==January 18, 2024==
 +
* OA showing different names highlighted when logging in the OA, now fixed on staging
 +
 
 +
 
 +
==January 11, 2024==
 +
* Duplicate function in OA was not working when using special characters. Valerio debugged and is now fixed.
 +
** Curators should make sure that, when pasting special characters, the duplicate function works
 +
* OA showing different names highlighted when logging in the OA, Valerio will debug and check what IP address he sees
 +
** If you want to bookmark an OA url for your datatype and user, log on once, and bookmark that page (separately for prod and dev)
 +
* Chris tested on staging and production the phenotype form and the data are still going to tazendra
 +
** Chris will check with Paulo. Once it is resolved we need to take everything that is on tazendra and put it on the cloud with different PGIDs
 +
** Raymond: simply set up forwarding at our end?
 +
* AI working group: Valerio is setting up a new account for open AI -paid membership for ChatGPT4. We can also use Microsoft Edge copilot (temporary?)
 +
* Chris getting ready to deploy a 7.0.0. public release - February 7th. Carol wanted to push out monthly releases. This will include WS291. For subsequent releases the next several releases will be WS 291 until WS292 is available.
 +
* Valerio would like to use an alliancegenome.org email address for the openAI account
 +
* New alliance drive: https://drive.google.com/drive/folders/0AFkMHZOEQxolUk9PVA
 +
** note: please move shared files that you own to new Alliance Google Drive.  Here is the link to the information that Chris Mungall sent:  For more instructions see the video and SOP here:https://agr-jira.atlassian.net/browse/SCRUM-925?focusedCommentId=40674
 +
* Alliance logo and 50 word description for TAGC> Wen will talk to the outreach WG
 +
* Name server. Manuel working on this, Daniela and Karen will reach out to him and let him know that down the road micropublication would like to use the name server API to generate IDs in bulk
 +
* Karen asking about some erroneous IDs used in the name server. Stavros says that this is not a big deal because the "reason" is not populating the name server
 +
* It would be good to be able to have a form to capture additional fields for strains and alleles (see meeting minutes August 31st 2023. https://wiki.wormbase.org/index.php/WormBase-Caltech_Weekly_Calls_2023#August_31st.2C_2023). This may happen after Manuel is done with the authentication.
 +
* Michael: primary flag with Alliance. Kimberly talked about this with the blue team. They will start bringing that over all papers and fix the remaining 271 items later.
 +
 
 +
==January 4, 2024==
 +
* ACKnowlegde pipeline help desk question:
 +
** Help Desk: Question about Author Curation to Knowledgebase (Zeng Wanxin) [Thu 12/14/2023 5:48 AM]
 +
* Citace upload, current deadline: Tuesday January 9th
 +
** All processes (dumps, etc.) will happen on the cloud machine
 +
** Curators need to deposit their files in the appropriate locations for Wen
 +
* Micropublication pipeline
 +
** Ticketing system confusion
 +
** Karen and Kimberly paper ID pipeline; may need sorting out of logistics

Latest revision as of 18:05, 11 April 2024

Previous Years

2009 Meetings

2011 Meetings

2012 Meetings

2013 Meetings

2014 Meetings

2015 Meetings

2016 Meetings

2017 Meetings

2018 Meetings

2019 Meetings

2020 Meetings

2021 Meetings

2022 Meetings

2023 Meetings

April 11th, 2024

  • Caltech WS293 ace files ready for the upload

April 4th, 2024

  • Continued discussion on sustainability
  • CZI, single cell RNAseq for Alliance -> anything happening will be few months down the road
    • Data is still going to SPELL and enrichment analysis
    • Peter Roy asking about expression profile of a condition and find similar expression profiles (SPELL like analysis) but SPELL cannot currently deal with scRNAseq data. Wen says it is possible (regarding each cell group as an experiment). Can try loading the into SPELL. Doe it improve the function of SPELL? Only 5-10 datasets. These data are a bit different from bulk RNAseq.
  • Textpresso: good to have a presentation for other MODs to show Textpresso capabilities? Yes. Maybe during sprint review
  • Michael's presentation on LLMs - Named Entity Recognition (NER)

March 14, 2024

TAGC debrief

February 22, 2024

NER with LLMs

  • Wrote scripts and configured an LLM for Named Entity Recognition. Trained an LLM on gene names and diseases. Works well so far (F1 ~ 98%, Accuracy ~ 99.9%)
  • Textpresso server is kaput. Services need to be transferred onto Alliance servers.
  • There are features on Textpresso, such as link to PDF, that are desirable to curators but should be blocked from public access.


February 15, 2024

Literature Migration to the Alliance ABC

Use Cases for Searches and Validation in the ABC (or, what are your common actions in the curation status form)?

Find papers with a high confidence NN classification for a given topic that have also been flagged positive by an author in a community curation pipeline and that haven’t been curated yet for that topic
  • Facet for topic
  • Facet for automatic assertion
    • neural network method
  • Facet for confidence level
    • High
  • Facet for manual assertion
    • author assertion
      • ACKnowledge method
    • professional biocurator assertion
      • curation tools method - NULL
Manually validate paper - topic flags without curating
  • Facet for topic
  • Facet for manual assertion
    • professional biocurator assertion
      • ABC - no data
View all topic and entity flags for a given paper and validate, if needed
  • Search ABC with paper identifier
  • Migrate to Topic and Entity Editor
  • View all associated data
  • Manually validate flags, if needed

PDF Storage

  • At the Alliance PDFs will be stored in Amazon s3
  • We are not planning to formally store back-up copies elsewhere
  • Is this okay with everyone?

February 8, 2024

  • TAGC
    • Prominent announcement on the Alliance home page?
  • Fixed login on dockerized system (dev). Can everybody test their forms?

February 1, 2024

  • Paul will ask Natalia to take care of pending reimbursements
  • Dockerized system slow pages (OA and FPKMMine). Will monitor these pages in the future. Will look for timeouts in the nginx logs.

January 25, 2024

Curator Info on Curation Forms

  • Saving curator info using cookies in dockerized forms. Can we deploy to prod?

ACKnowledge Author Request - WBPaper00066091

  • I am more than willing to assist; however, the task exceeds the capabilities of the normal flagging process.
  • The paper conducts an analysis of natural variations within 48 wild isolates. To enhance the reliability of the variant set, I utilized the latest variant calling methods along with a custom filtering approach. The resulting dataset comprises 1,957,683 unique variants identified using Clair3. Additionally, Sniffles2 was used to identify indels of >30 bp, which numbered in the thousands to tens of thousands for most wild isolates. It is worth noting that variants identified with Sniffles2 have less reliable nucleotide positions in the genome.
  • I am reaching out to inquire whether WormBase would be interested in incorporating this dataset. An argument in favor is the higher quality of my data. However, I am mindful of the potential substantial effort involved for WormBase, and it is unclear whether this aligns with your priorities.
  • Should WormBase decide to use my variant data set, I am more than willing to offer my assistance.

Update on NN Classification via the Alliance

  • Use of primary/not primary/not designated flag to filter papers
  • Secondary filter on papers with at least C. elegans as species
  • Finalize sources (i.e. evidence) for entity and topic tags on papers
  • Next NN clasification scheduled for ~March
  • We decided to process all papers (even non-elegans species) and have filters on species after processing.
  • NNC html pages will show NNC values together with species.
  • Show all C. elegans papers first and other species in a separate bin.

Travel Reimbursements

  • Still waiting on October travel reimbursement (Kimberly)
  • Still waiting on September and October travel reimbursements (Wen)

UniProt

  • Jae found some genes without uniProt IDs, but the genes are there on uniProt but without WBGene IDs.
  • Wen reached to Stavros and Chris to investigate WormBase and AGR angles.
  • Stavros escalates the issue on Hinxton Standup.
  • Mark checks Build scripts and WS291 results. After that, he contacted UniProt and he's working with them to figure this out.

January 18, 2024

  • OA showing different names highlighted when logging in the OA, now fixed on staging


January 11, 2024

  • Duplicate function in OA was not working when using special characters. Valerio debugged and is now fixed.
    • Curators should make sure that, when pasting special characters, the duplicate function works
  • OA showing different names highlighted when logging in the OA, Valerio will debug and check what IP address he sees
    • If you want to bookmark an OA url for your datatype and user, log on once, and bookmark that page (separately for prod and dev)
  • Chris tested on staging and production the phenotype form and the data are still going to tazendra
    • Chris will check with Paulo. Once it is resolved we need to take everything that is on tazendra and put it on the cloud with different PGIDs
    • Raymond: simply set up forwarding at our end?
  • AI working group: Valerio is setting up a new account for open AI -paid membership for ChatGPT4. We can also use Microsoft Edge copilot (temporary?)
  • Chris getting ready to deploy a 7.0.0. public release - February 7th. Carol wanted to push out monthly releases. This will include WS291. For subsequent releases the next several releases will be WS 291 until WS292 is available.
  • Valerio would like to use an alliancegenome.org email address for the openAI account
  • New alliance drive: https://drive.google.com/drive/folders/0AFkMHZOEQxolUk9PVA
  • Alliance logo and 50 word description for TAGC> Wen will talk to the outreach WG
  • Name server. Manuel working on this, Daniela and Karen will reach out to him and let him know that down the road micropublication would like to use the name server API to generate IDs in bulk
  • Karen asking about some erroneous IDs used in the name server. Stavros says that this is not a big deal because the "reason" is not populating the name server
  • It would be good to be able to have a form to capture additional fields for strains and alleles (see meeting minutes August 31st 2023. https://wiki.wormbase.org/index.php/WormBase-Caltech_Weekly_Calls_2023#August_31st.2C_2023). This may happen after Manuel is done with the authentication.
  • Michael: primary flag with Alliance. Kimberly talked about this with the blue team. They will start bringing that over all papers and fix the remaining 271 items later.

January 4, 2024

  • ACKnowlegde pipeline help desk question:
    • Help Desk: Question about Author Curation to Knowledgebase (Zeng Wanxin) [Thu 12/14/2023 5:48 AM]
  • Citace upload, current deadline: Tuesday January 9th
    • All processes (dumps, etc.) will happen on the cloud machine
    • Curators need to deposit their files in the appropriate locations for Wen
  • Micropublication pipeline
    • Ticketing system confusion
    • Karen and Kimberly paper ID pipeline; may need sorting out of logistics
Retrieved from "https://wiki.wormbase.org/index.php?title=WormBase-Caltech_Weekly_Calls&oldid=38411"