wiki:PreviousCurationMeetings

Version 73 (modified by vw253, 6 years ago) (diff)

--

Meetings held

6 Sept. 2011

  • Annotation extensions (with Chris, Emily, Jane; morning)
    • relations
      • direct and indirect regulation - has_regulation_target, plus child term(s) to be added (see what Chris has put in file)
      • when to use has_substrate - with MF catalytic activity terms
      • upstream & downstream
      • negating annotation extensions
  • Gene expression controlled curation
    • current capture of expression (some needs to migrate to annotation_extension)
    • how should we do this in future?
  • Phenotype modelling (with Kim & Mark; afternoon)
  • WT phenotypes

3/4 October 2011

  • As part of planning meeting Midori went through phenotype model for Chad (link to planning meeting minutes)
  • As follow on from phenotype modelling we drew up a draft workflow for capturing data in the curation tool (We will all work through this when curating phenotypes)
  • Midori went over basics of using subversion, esp. resolving conflicts
  • Bio-Curator meeting abstracts (Nov 30).http://pir.georgetown.edu/biocuration2012.html/instructions.html hopefully Antonia can attend
  • General curation work

17/18 October 2011

  • Antonia's list

  • Regulation
    • Direct vs indirect
    • regulation of cell cycle and signal transduction
    • start and end of a process

3/4 November 2011

  • Identified and co-curated papers for undergraduate workshop

14 November 2011

  • round robin, next 2 weeks priorities etc

  • GO meeting debrief
  • Had a discussion about annotation extension When to pre-compose in ontology (during etc)
  • Also discussed phenotype curation, probably only need condition extension initially
  • Discussed how location required for process is equivalent to process dependent on localization to componet
  • need to migrate all required_for attached to components to process terms and move all dependent on to children
  • PomBase release pending, discussion about a few thing required for release, and revised release date to 28th

15 November 2011

2 December 2011

  • round robin, next 2 weeks priorities etc
  • Midori OBO-Edit basics
    • loading viewing browsing searching etc (need to be able to view relations ontology)
  • Posters for biocuration

23-27 January 2012

Helpdesk

  • RT versus jira (For a while we will probably stick with RT)
    • RT is now be configured so we see all responses.
    • Val demo how to configure dash to show all open/ closed items / assign itmes DONE
    • Val will encourage people to submit to helpdesk, or will summarize and post queries later
    • Note make sure any feedback received is appended to closed items as comments.
  • Proposed system:
    • Antonia monitors the helpdesk
    • Antonia decides if she can't answer the query who would be able to best answer (probably me at first for most legacy or data related questions, and Mark for inclusion of HTP datasets, Midori for PomBase website queries (including queries which will raise new jira tracker issues or feature requests), Antonia for requests to curate papers), but we can always reassign if they go to the wrong person...)
    • Once a query is answered Antonia will draft an FAQ item (if relevant) from the response, and send to Midori for editing and including in the PomBase FAQ
    • Hopefully after a year or so we will have answered most of the common queries and so increasing Antonia can increasingly point users directly to the FAQ.

Curation Tool

  • Getting feedback from authors on completed papers Val will write an email to send out with curation sessions We should keep a file of people who have been sent completed sessions as the next paper they should curate themselves

To discuss with Kim

  • Next things for Kim to work on (rough order) Curation tool

  • need to thrash out allele descriptions
  • Val needs to check that the proposal to change the tool fits with the phenotype format/model
  • 4. Annotation extensions
    • I.e which extensions need to be available for i) each ontology ii) each term
    • ii) annotation transfer of extensions and qualifier
  • 5. New datatypes, expression (dependent on outcome of discussion below), complementation (probably we will split out complementation as a curation type in the tool), protein sequence features (subset of SO)
    • GO through high priority SF items (1 & 2 above)

  • New features
    • See SF tracker

Mark

All items from discussion with mark added to relelvant jira tracker Other Action items

  • Download file options (val to document)
  • replacing current ftp site and download files (val to document) DONE

  • How to capture dephosphorylated/dephosphorylation DONE
  • pombelist FAQ proposal (putting pombelist "experimental" items into Susan F's FAQ)

6-7 February 2012

  • went through Antonia extensions (kim generated file with links to sessions)

Antonia will revise existing extensions to fit parser

  • extensions on phenotypes generally

Discussed that it is good practice to add targets to both phenotypes and GO terms, and that more generally extensions will be required for phenotypes, initially these wll need to be done by curators.

  • later we will be able to make links between phenotypes and GO terms
  • relations used in extensions will be reported by Kim during loading and the we can refine
  • Val went through some localization examples , e.g PMID:20298435
    • should the localization only be direct transport, anchoring, chaperoning? what about complex assembly, or part of a complex
    • decided localizations should be captured with GO if they were clearly "localisation" (i.e A- required for B). If possibly "complex dependent" i.e A required for B AND B required for A, should be phenotype only)

  • Using OBO-Edit, part 2: editing & saving
    • make a "condition" ontology

5-6 March 2012

Phenotype ontology

  • Should we keep some, all, or none of the "abnormal protein localization" terms? If some, which? Also see SF 3485679 (Midori 2012-02-29)
    • Decision as of 2012-03-05: keep all (unless a specific reason to obsolete a term crops up), because we'll need 'delayed' and therefore also a generic abnormal-process parent; pepper ontology with comments galore to explain which terms to use when.
    • Midori to try to sort out ontology structure with advice from George & Chris. It'll all work out somehow ...
  • Different terms if the mutated protein is mislocalized, or a target of the mutated protein is mislocalized? (Antonia 2012-03-01)
    • Decision as of 2012-03-05: use the same phenotype ontology term, and use an extension to identify which protein is mislocalized (always include an extension, to show that we know which protein was assayed).
  • Also, what extensions to use?
  • Discussed order of sections on Gene Summary pages/ NOT annotation submitted jira items
  • went through possible slides for Antonia's Biocurator presentation
  • Chased up follow up for user testing
  • went through protocol for assigning products (precedence now on wiki)

27th March 2012

  • Antonia's talk preparation
  • User testing discussion

17th April 2012

  • Skype with Kim to go through allele phenotype workflow

01 May 2012

EBI user testing, an planning meeting

14th May 2012

https://sourceforge.net/tracker/?func=detail&aid=3526555&group_id=65526&atid=2525791

  • Go through new BioGRID documentation for genetic interactions, related to our items below.
    • capturing or inferring recessivity/dominence (and capturing heterozygosity or homozygosity in a diploid)
    • Will discuss when we have real life examples

  • Which relationship to use to represent target for normal protein localization..... (Midori suggested "assayed using" ...I'm happy to start with that...DONE,

23rd July 2012

  • Phenotype assumptions: do we want some 'default' conditions that are true for most experiments, unless otherwise stated? See for instance the terms decreased aerobic cell growth on glucose carbon source / slow cell growth / etc

DONE (in part) Decided to make some assumptions for "standard" conditions, and request terms for non standard conditions (if standard condition terms are requested, they can be added as synonymns and liberally commented) https://sourceforge.net/tracker/?func=detail&aid=3547584&group_id=65526&atid=2096431

  • Decide workflow for annotation transfer for phenotypes in the curation tool

DONE https://sourceforge.net/tracker/?func=detail&aid=3547608&group_id=65526&atid=2525791

  • workflow for curation tool, lab/person

DONE https://sourceforge.net/tracker/?func=detail&aid=3547630&group_id=65526&atid=2525791

  • Discussion about annotation specificity
    • curators collect examples of where they wondered whether they were going "too specific"
    • Curators to co-curate Espenshade papers to get a feel what degree of specificity we want to capture

DONE made some progress with this, basically in some cases for "response to" we will use extensions AND concurrent annotations so we don't have explosions of some conserved areas of the ontology which are likely to be otherwise conserved (especially when the same pathway can be triggered in response to various insults or depletions)

Also looked at a specific paper and decided that some compound terms were required to capture annotation specificity, AND there were a couple of possible issues with this existing part of the ontology (around SRBP signalling)

21st August 2012

  • Discussed with Mark:
  • Big item: display of annotation extensions (GO, FYPO), 'with' column (GO), alleles (FYPO), etc.
  • Other items:
    • set up regular meetings - one day of every other curator meeting, i.e. once every 4 weeks
    • gene page aesthetics (http://www.ebi.ac.uk/panda/jira/browse/PB-719)
    • improving notification about updates to "module", "biomarts", etc.
    • note for later meeting: collapsing display for annotations with lots of extensions

17th September 2012

  • Make a mapping of relations to display on pages
    • Draft now in Dropbox/pombase/PomBase_website/extension_relation_display_text.txt (2012-09-17)

5/6 th November 2012

  • N-matrix
  • skype with Kim, discussed multi gene phenotype worksflow in curation tool (see slides in dropbox)

7th Jan

  • Discussed how simple and site-wide PomBase search should behave (midori will put on Jira tracker)
  • Discuss ad/disad of abolished enzyme activity as post composed...new_relation(GO:xxxxx) Discussed, and no time saving.
  • Gene expression data: we have a (sort of) temp solution for few compounds many genes. But no solution for 1000s of compounds + 1 gene. What to do with these? Eg. PMID:22840777 We discussed this and decided that what Antonia did here was adequate
  • Gene expression data: How to rejig these. Currently a bit suboptimal (i.e. veg growth, and present vs uchanged vs increased etc). Kim needs to know what is allowed and what is not. -> postponed, not ideal but works for now
  • Phenotype ontology:
    • Decide how to deal with glucose repression SF item Agreed see SF item
  • Check over 'sensitive to high temperature' and child terms 'decreased growth at high temperature' and 'inviable at high temperature'. Make sure current arrangement makes sense. (Also see 'Pre- vs post-composing terms' below and SF 3588912) Agreed to proceed with only inviability and decreased growth pre-composed, as people will search on "temperature sensitive" other terms will use "at high temperature" as a condition
  • Pre- vs post-composing terms
    • background (older comments):
      • We've decided to make some phenotype terms obsolete and use conditions to capture temperature, carbon source, nitrogen source, etc. Last call to review the affected terms -- are there any that we should keep after all?
        • normal growth at high temperature FYPO:0000674
        • viable at high temperature FYPO:0000282
        • sensitive to cold FYPO:0000080
        • sensitive to high osmolarity FYPO:0000081
        • sensitive to high temperature FYPO:0000082
        • slow growth at high temperature FYPO:0000247
        • normal cell growth on ammonia nitrogen source FYPO:0000242
        • normal cell growth on proline nitrogen source FYPO:0000243
        • slow cell growth on ammonia nitrogen source FYPO:0000249
        • slow cell growth on proline nitrogen source FYPO:0000250
        • slow cell growth on galactose carbon source FYPO:0000251
        • slow cell growth on glycerol/ethanol carbon source FYPO:0000442
        • increased cell growth rate on glucose carbon source FYPO:0000637
        • slow cell growth on glycerol carbon source FYPO:0000684
    • update 2012-03-09: I'm now thinking of keeping just heat sensitive (FYPO:82) and cold sensitive (FYPO:80) as a practical matter -- lots of community curators/users will search for them! The rest could probably still go, converted to parent + condition(s).
    • update 2012-06-12 (based on discussion in May): actually still thinking about this. It's a "when to pre- or post-compose terms" question, analogous to GO terms + extensions. Thoughts in progress noted here.

We decided we are happy with how we are doing this right now

Date ?

  • Abstracts (done)
  • annotation intersects, rules update (Done)
    • Val will make poster/slides/text

Monday 22 April

  • Antonia will paste in her "protein glycosylation" so far into Matrix results (DONE)
  • check that the qualitative/quantitative expression stuff is all documented and working OK (need list of allowed qualifiers etc) (DONE OR IN PROGRESS)
  • Call with Kim to chat about Curation tool :
    • request to curate a paper e-mail, revisions (IN PROGRESS WE 3 VW,MH,AL, will do thorough testing once next batch of changes are made to test)
  • PomBase User Documentation Drafts (see wiki, mainly done, we should check these through for details (i.e where TMHMMs from) and move onto jira for mark to link to page sections, and announce) (IN PROGRESS, MOST POMBASE USER DOCUMENTATION IS IN PLACE, NEED TO LINK FROM SPECIFIC PAGE SECTIONS TO HELP, and ANNOUNCE)
  • make some FYPO terms unavailable for direct annotation. There are lots of terms where it should be possible to make a more specific phenotype annotation (i.e swollen, elongated, multinucleate, aseptate etc), where it is possible that we did not know this when we made the original annotation. We should slowly look through these and move down. An example but not a great one: https://sourceforge.net/p/pombase/curation-tasks/181/...we we should make the curation tool force people to annotate more specifically for these. For example they should never be able to annotate to a term which does not specify vegetative/ vs G0/ meiotic etc.... (DISCUSSED, AS WE NOTICE TERMS WE CAN ADD THIS SUBSET, AND "DO NOT ANNOTATE" COMMENT. ASKED KIM TO ADD A WARNING TO LOGS)

  • "response to" phenotype only?
    • Antonia: when to use cellular response to terms (I know they should not be used for expression data and not to use 'toxin' or 'drug'. But what if there is a sensitivity/growth assay? they are sort of informative are they not - I just want a reminder)

(DISCUSSED, WE WILL NOT USUALLY USE "RESPONSE TO" Terms for GO TERM ANNOTATIONS ANY MORE, BUT WE WILL USE TO MAKE PHENOTYPE ANNOTATIONS AND USE "IN RESPONSE TO" IN EXTENSIONS)

  • How to alert users to scope of terms with 'during vegetative growth' in synonyms and xps, but not the term names? Could rename all, or add to defs, or add comments. (IN PROGRESS)
    • i) Midori will add "vegetative growth to definitions"
    • ii) We will ask Kim if it is possible to have a "group" to use a different synonym_type_def and if we did this would it be possible to display this term name as a default for these terms in the curation tool
    • iii) if not we will add to term names (will be a bit clunky on gene pages)
    (example processes localisation to centromere/ spindle pole body duplication)

( DISCUSSED THIS SHOULD BE FINE AS SYNONYMS ARE NOW ADDED, AND EXACT SYNONYMS WILL BE DISPLAYED BY DEFAULT FOR PHENOTYPE)

  • Discussion about viability/ inviability cell or population phenotype (WILL FOLLOW UP WITH SOME QUEIES TO CHECK ANNOTATION AFTER V 35)

Discussed the following "Reminders"

  • keep a look out for these: http://curation.pombase.org/pombe/curs/cc872132c252909e look for others in response waiting file I changed the evidence for the IGI annotations to ISO because it wasn't really shown in this paper that it is involved in those processes, it only rescued a growth defect in cerevisiae ortholog mutant
  • check all community curation flags set on approved papers

For any papers you have sent out, can you check that the "community curatable" flag is set in the triag (also that they have triage status "curatable", some appear to have had the papers added, but the status wasn't set (probably because it was added directly rather than by the automated paper load).

  • Val: I might have put some things through to abnormal chromosome/chromatin organization that should probably be morphology...could you have a look out for them?

(we should do queries on this for V 34 and V35 to check) https://sourceforge.net/p/pombase/fission-yeast-phenotype/647/

  • localization dependency (many should move to phenotype only during reannotation)

13 May 2013

  • pombe2013 meeting plans
    • some sort of presence at opening reception, to play up community curation launch
    • lunchtime demo sessions
      • Tuesday 25th, 12:25-13:15 Community Curation Demo
        • Antonia - demo of the interface
        • Midori - more on phenotypes -- normal/ abnormal; cell vs population, differences between phenotypes and GO biological process annotations (direct/indirect), localization dependency, etc.
      • Thursday 27th, 12:25-13:15 PomBase Demo
        • Val - Gene Page walkthrough highlighting features and Advanced search (note "list upload" feature)
        • Mark- Ensembl browser demo (specifically configuring tracks and sequence region download)
    • sod t-shirts; go for giveaway trinkets instead (looking into "keyring dynamo torch", or can go for battery-powered if budget won't stretch to the wind-up kind)
    • banner in progress
  • should we obsolete the at high temperature growth phenotypes and recapture as conditions?
    • Maybe the obsolete terms could show this as a comment to make it easier for curators/community? (or are obsolete terms not accessible in the curation tool, I don't think I have any seen any although they would be useful for referring to an extant term) ....
    • see email thread "when to use high temperature" phenotypes
    • We decided (again) to carry on with the terms we have. It was useful to discuss again, though, to keep our heads round what we're doing and why, and to confirm that we don't want to change practices at present.

  • Val will make a draft PomBase? flyer for Pombe 2013

Antonia will re-curate this paper, and then we will decide which BP can really be made from these phenotypes

Val will do similar for some cell cycle phenotypes-> biological processes (elongated) after discussion with Jacky Midori needs the feedback from this exercise for her part of pombe 2013

Last few meetings 2013

Done 21st/22 Jan 2014

  • Discussed BIoGRID issue reported by Kathy yesterday, and by Henry previously
    • see my response waiting e-mail folder
      • waiting to hear back from BioGRID
  • brainstormed mockup of graphical display of modifications and mutations in sequence context
    • Done a first pass, Val (or somebody) will mock up, see Ensembl variation image
  • look at these process annotations http://www.pombase.org/spombe/result/SPAC1783.07c#fission_yeast_phenotype (compare function/ process, and consider in response to)
  • Val to document that annotations should be made preferentially on function with regulation targets and conditions.
  • Obsolete annotation in biological process can be removed AI Val tidy pap1 TF as discussed at curator meeting / promoter DNA binding, it has the promoter and tf regulated. Always represent on MF if possible

Canto

  • text, for whether a paper contains curatable gene or not https://sourceforge.net/p/pombase/curation-tool/624/ (is this OK?) (YES, OK)
  • text, Add allele and evidence misleading "add more alleles using this term" (non issue)
  • text, do you want to transfer evidence or condition (https://sourceforge.net/p/pombase/curation-tool/685/)
  • evaluating curtain tool, where do people expected to go (gene page or summary page) after annotation completion? - if I have annotated a single gene I expect to stay on that page. I f I have transferred annotations I feel as though I should go back to the summary page. Or is that odd? (non issue)

FYPO

  • cut phenotype (DONE new ticket raised to correct existing term, will need to obsolete and reannotate)

Physical Interactions

  • Antonia - Should we add GO/protein binding terms for enzyme/substrate combinations?
    • Antonia should we use the BioGRID evidence code "biochemical activity" which we don't use at the moment.
      • No but can capture high confidence with GO protein binding
  • Discussed when to capture physical interactions in GO using 'protein binding' in general.
    • Only for direct, high confidence interactions. I.e within complex, or modification/target

PomBase gene page overhaul

  • Decided how we would like to display annotations to streamline/ hide redundancy
  • changes to data layout and sorting on gene page, user configuration Val to mock up and document

Feb. 2014

Consistent labels for track displays

May 8th/9th 2014

  • Elena Hidalgo's feeedback,
    • We should go through this feedback and work out if the is a way to improve the capture of the "restored growth"

  • decided to:
    • Antonia will complete session and approve along the lines of Elena's requests
    • will prioritise for double mutant annotation
    • change term names to remove 'cell population' in these cases https://sourceforge.net/p/pombase/fission-yeast-phenotype/1485/
    • auxtrophic for sulphur contining https://sourceforge.net/p/pombase/fission-yeast-phenotype/1485/
    • note that auxotrophic is a phenotype which does not encompass viability information becasue it is conditionally inviable and so we make an additional annotation to "viable or inviable population terms under the given conditions, (i.e inviable without methionine, normal growth when methionine added)
      • AI get back to elena when double mutant curation possible

secretion vs. exocytosis

"If proteins are secreted only by exocytosis in fission yeast, I can add synonyms to the FYPO terms, even though the corresponding synonyms wouldn't work for GO (nor for the generic cell phenotype uber-ontology)" (e-mail 'mug33 checks 2')

  • AI: be aware that if you annotate in GO to exocytosis, you are not annotation to protein secretion, or intracellular protein transport. Are new child terms required to capture these parenatages?

explain situation where I would use "sensitive to heat"

binucleate/ multinucleate

  • AI postponed until follow up with Jacky

cut remapping

  • AI Val, remap genes mapped to old cut to new cut where I know, and then pass on to Antonia if any remaining

localization/transport/targetting

during mitosis (process)

difference between inviable spore and abolished spre germination, term merge

  • Merge 'inviable spore' (or inviable spore population?) and 'abolished spore germination' - almost certainly a good idea, since there isn't really any practical way to gauge spore viability other than to see if it germinates. But which name and logical definition should we keep? (SFs 592, 785 and any linked)
  • also consider cell vs. population angle for sporulation occurrence/frequency terms (SF 1319)
  • obsolete FYPO:0000279 because azygotic sporulation isn't abnormal; it happens under specific growth/nutrient conditions (some info in SF 1431) - done

GO protein binding

GO meiosis term name changes

  • A heads up meiosis (mieotic nuclear division) vs meiosis (meiotic cell cycle)
  • Note the difference, especially in extensions during (meiotic cell cycle)

when to request a new GO term

We think we know when but

  • AI, discuss if necessary by e-mail as and when we have specific examples

has_direct_input discussion

  • How many has_direct_input do we have on processes- Could the annotation be represented more specifically as a molecular function https://sourceforge.net/p/pombase/curation-tasks/368/ i.e It isn't necessarily wrong, but it isn't the most specific possible annotation if it is has_direct_input and it isn't on the MF term
    • Discuss use of has_direct_input (a.k.a. has_substrate) v.s. has_indirect_input v. the parent has_input
    • At its simplest I can't think of an example where we would not use has_direct _input with a MF term. I would use has_input or has_indirect input for upstream regulators (i.e BP regulation of enzyme or process terms).

We all agree that we haven't come across any exceptions yet, and we annotate more specifically to the MF

drectionality of protein domain in protein binding annotation extension

  • Query from Antonia Subject d384a52dcc0dc8a1 about Should the substrate be the domain or the gene product?
  • AI check how many protein domains in extensions, and their types
  • AI ask go to extend scope of occurs_at to include domains (includes SO protein features)

Do not annotate subset

  • Do not annotate subset for FYPO go in the ontology (need reporting in logs, there is a jira ticket about this)
  • GO has general do not annotate subset
  • We have our own GO do not annotate subset
  • Midori made a file in /pombe-embl/mini_ontologies/GOterms excluded form Pombase
  • AI add to Kims ticket i) report annotatin in chado logs ii) make canto not search/display these terms if they are leaf nodes or does not have descendents that are avaiable for annotation. For others just flag a commnet that this is not for direct annotaton. note for Kim, do not propagate up or down. Note for us add terms, consider protein import/export
  • AI Do not annotate subset...Keep an eye out for terms which could be included (not in logs yet)

has_part discussion

  • Discuss has_part in relation to annotation propagation in Canto
  • SF 566 use has_part relationships for ontology term drill-down (related thread is 'using has_part for ontology drill down' and NTR: srebp catabolism for an illustration Q are phenotype has_parts visible in the curation tool?
  • Discuss when annotations should inherit over FYPO (or alternative solutions) https://sourceforge.net/p/pombase/curation-tasks/188/ e.g. has_output between process and structure phenotypes, e.g. abolished actomyosin contractile ring assembly has_output contractile ring absent
  • long discussion about has_part
  • summary
    • AI Curators need to keep a look out and make concurrent annotation for things which are only related by has part, or request more specific term, where appropriate
    • AI propagation over has_part is appropriate for phenotypes (searching) and a jira ticket will be raised to make this happen

Sept 17th 2014

display of phenotypes with 2 targets

  • for phenotypes with two "targets" (e.g. protein binding terms) the affecting relationship is displayed twice on two rows. perhaps a bit cryptic? Change to something else? For instance:
    "affecting cdc2 and suc1" instead of

"affecting cdc2
affecting suc1"

DONE

Issues related to branching and T-shaped phenotype

https://sourceforge.net/p/pombase/fission-yeast-phenotype/1696 https://sourceforge.net/p/pombase/curation-tasks/427/

Grouping for antioxidants

  • representing core stress response genes and "response to stress", specifically

i) stress signalling ii) antioxidants etc (thread 'Annotation extension paper' this came up in AE pper egs)

will be resolved by https://sourceforge.net/p/geneontology/ontology-requests/11063/

Regulation question

  • we decided this was regulation and assigned to Antonia

Transcript IDs

https://sourceforge.net/p/pombase/curation-tasks/385/