PAINT 2014 Summer Jamboree Kickoff Meeting Logistics

From GO Wiki
Jump to: navigation, search

Dates

  • July 28-31, 2014

Hotels

Five hotels are within a 15-30 minute walk to the meeting venue. Ordered below by price per night, lowest to highest.

  1. Hotel California
  2. Cardinal Hotel
  3. Stanford Terrace Inn
  4. Sheraton Palo Alto Hotel
  5. Westin Palo Alto

Venue

3165 Porter Dr, Palo Alto, CA
Room 1118 (downstairs)

Porter Drive is not on campus, it is off of Page Mill Road.

There is plenty of free parking in the front and rear of the building.

Agenda

Monday, July 28th

Schedule

9:00am - 12:00pm - Introduction

12:00-1:30pm - Lunch

1:30pm - 5:30pm - Jamboree

Discussion and Action Items

Discussion of specific Panther families

  • PTHR13348 - Li brought up a question for PTHR13348 where the Annotation tab is showing a single ID in the with field for the "protein binding" annotation even though there are actually three ID's listed in the with field in MGI's annotation. Suzi explained that this is a parsing issue that the with field is so irregular that when there are more than one with ID's for a single annotated gene with the same reference, only one is displayed. This is not trivial to change. We agreed to accept this as an issue to be aware of as is.
  • PTHR24073 - Karen brought up a question about this family, which is the Rab GTPase superfamily. The family is huge (>2700 sequences) with a duplication node near the top. The majority of the nodes immediately below the duplication node clearly represent major classes of Rab proteins, e.g. Rab2's, Rab21's, Ift27's, Ift22's. Having scanned through the tree and annotated genes, in some clades, there is experimental evidence for "GTPase activity". However, in others, there is only evidence for "GTP binding" and/or "GDP binding". Karen asked about how to handle this family, specifically the fact that it is so big that she is confident about only a few subclades, not about all of them.
    We agreed that for huge families like these, we can do an "in progress" commit, using a specific message in the SVN commit log.
    We discussed what is the simplest explanation for GTPase activity in many of these clades, though it is not shown in all of them. Paul urged strongly to go ahead and propagate the "GTPase activity" MF term to the root of the tree. While this may be inappropriate for some subclades, let's go ahead and do it, and as Karen and/or other curators work through various subclades, and we get feedback from MOD and other curators, we can block propagation of this term to appropriate clades as appropriate.
    ACTION ITEM (Karen) - Update the PAINT User Guide with info about committing an "In Progress" family (both SVN details and philosophy), include link to web view of SVN log messages, and send out an email to let everyone know.
    ACTION ITEM (Suzi/Paul) - Update/modify the upcoming Jenkins report so that it will include previous commit messages, so that curators will be able to see the "in progress" log note in the web viewer for the SVN repository.
  • PTHR11559 - Huaiyu brought up a question about general function vs. specific function. The family is a carboxylesterase family. One clade has cholinesterase genes. Within that clade, there is another clade that are acetylcholinesterase genes, which have more specific molecular function. One cholinesterase gene is annotated with acetylcholinesterase activity. However, the gene also has other cholinesterase activity. The tree shows clearly that the acetylcholinesterase specificity is gained during evolution.
    We agreed that this is something that we need to discuss with the annotation group.
    The specificity is more obvious in context of the tree.

Other Action Items

  • DONE - ACTION ITEM (Suzi) - In PAINT, clicking on links within the Annotations tab should NOT recenter, but clicking on a term in the Annotation matrix should recenter to the LCA
  • ACTION ITEM (Suzi) - send out contact info for people to let know about when you see a bunch of partial sequences

Tuesday, July 29th

Schedule

9:00pm - 10:00pm - Daily kickoff discussion

10:00am - 12:00pm - Jamboree

12:00-1:30pm - Lunch

1:30pm - 5:30pm - Jamboree

Discussion and Action Items

  • Propagation of qualifiers - In PTHR23074 (AAA ATPase), Pascale brought up a question about this ATPase family and whether it was appropriate to propagate the "contributes_to" qualifier when this subunit has an ATPase domain and most likely actually has ATPase activity on its own. This brought up the fact that the wording of the pop-up box is potentially confusing, especially to new users or those who rarely use the qualifier pop-up box.
    DONE - ACTION ITEM (Suzi) - come up with clearer wording for the popup - that the yes/no is asking only about whether or not to propagate the qualifier, the term will be propagated either way
Do you want to also propagate the qualifier
contributes_to (from HUMAN_XXXXX)
  • Logic of how far up to propagate - Also in PTHR23074 (AAA ATPase), Pascale asked why the MF term "ATPase activity" was propagated only up to the top of the eukaryotes, but not to the root of the tree including the Archaea. Pascale explained that when the MSA and domain conservation (checked using UniProt pages) is good, she feels confident that she can propagate all the way to the top so that the Archaeal sequences also get IBA annotations.
    • DONE - ACTION ITEM (Moni) - finished annotation of this tree (partially done for presentation in June)
      DONE - ACTION ITEM (Suzi/Huaiyu) - Would be great if MSA would adjust when nodes were pruned. Huaiyu needs to tell Suzi which info in the tree file indicates a horizontal transfer node.
  • Horizontal transfer events - These are now shown in PAINT (version 9).
    • PTHR23152 - In this specific family, this gene was transferred to eukaryotes as part of mitochondrial engulfment. Be aware that this is an accurate representation of the evolutionary relationship, and make sure to propagate terms like "mitochondria" only to the eukjaryotic seqs, not the bacterial ones.
    • PTHR16943 - PantherDB.org shows horizontal transfers
      This is a complicated tree. PrpD annots can go up to top of that clade. IRG1 annots can go up to Craniata-Cephalochordata node.
      DONE - ACTION ITEM (Suzi) - linkouts to UniProt IDs no longer present in 1.7
      DONE - ACTION ITEM (Suzi) - PAINT needs to show horizontal transfer nodes
  • Taxon checks - It was asked if these are active in PAINT?
    These are definitely on during propagation. If you see a problem during propagation, the first thing to check is if the taxon constraints in GO are correct.
  • Questions for Tanya
    • Should the methyl cycle (aka SAM cycle) be part_of "methylation", or should it be the other way around? Alternatively, should there be any relationship between the two?
  After discussion with other GO eds, we will make the new relationship: SAM cycle has_part 'S-adenosylmethionine-dependent 
  methyltransferase activity'. The latter term is transitively linked to 'methylation' over par_of by MF-BP links.
    • U7 snRNP: Should this be a child of "nucleus" and "cytoplasm"
  No.  Not part_of.  Because u7 snRNPs are found in both nucleus and cytoplasm. Could do 'cytoplasmic U snRNP body' has_part 'U7 snRNP' 
  if the cytoplasmic body always has U7 snRNP.  (There is no 'nuclear U snRNP body', could add that.) 
  A part of B means whenever A exists, it is part of B.
  part_of documentation
  A has_part B means that whenever A exists, B will always exist as a part of A.
 has_part documentation

Data = U7 snRNAs and dLsm10/dLsm11 (U7 snRNP core proteins) are found in nucleus *and* cytoplasm (immunolocalization, FISH, YFP fusion)

  • Additional PAINT requests for Suzi
    DONE - ACTION ITEM (Suzi) - In the With field in the Annotations tab, please hyperlink when the ID is from UniProt.
    DONE - ACTION ITEM (Suzi) - Is it possible include PMID in the find searches to locate gene(s) that it the paper is associated with?
    DONE - ACTION ITEM (Suzi) - In the "Update MSA Parameters" menu, please change the color of the text from white to black so that it shows up in the colored background.
    DONE - ACTION ITEM (Suzi) - Remove TAS from Reactome
    DONE - ACTION ITEM (Suzi) - Bug: "NOT" annotation of child term removes visualization of parent terms, even when parent terms have been propagated to an ancestral node

Wednesday, July 30th

Schedule

9:00pm - 10:00pm - Daily kickoff discussion

10:00am - 12:00pm - Jamboree

12:00-1:30pm - Lunch

1:30pm - 5:30pm - Jamboree

Discussion and Action Items

  • Question for Tanya

Should mRNA cap binding complex (5845) be a parent of nuclear mRNA cap binding complex (5846)?

  I don't think so.  The def of 5846 states: "A conserved heterodimeric protein complex that binds to the 5' terminal cap structure 
  m7G(5')ppp(5')N of nascent eukaryotic RNA polymerase II transcripts such as pre-mRNA and U snRNA."  Because the transcripts 
  could be U snRNAs, making 'mRNA cap binding complex' a parent of 5846 would be incorrect.
  • Additional PAINT requests for Suzi
    1. CANNOT BE DUPLICATED In some trees when there is a duplication node at the root, once you have propagated annotations to the root, you cannot view the annotations that were propagated to the root, and thus cannot delete them (Pascale & Li - PTHR10546). Li and Suzi tried to duplicate the bug, but were unable to.
    2. DONE PTHR14699 was just annotated yesterday (by Karen), but when it is reopened in PAINT 1.8 today, the notes section gives a warning message indicating that all of the annotations have been removed: "already has descendents with more general annotations". All of the propagated annotations were direct annotations to the exact term propagated.
    3. DONE Bug: "NOT" annotation of child term removes visualization of annotation of parent terms in the same clade, even when parent term(s) were annotated to ancestral node.
    4. LOW PRIORITY In PAINT 1.9 and in all earlier versions tested, some annotations are shown twice in the "Annotations" tab (PTHR11242), deemed low priority since no one has noticed this until now
    5. DONE The Open Menu no longer automatically selects the *.paint file and grays out the unopenable files. Thus now you have to manually select the *.paint file in order to open the right one.
    6. CANNOT BE DUPLICATED, It's back (maybe??)! If you load a new family in, after you have already annotated a family (or loaded an annotated family into memory), when the new family loads it, all of the notes from the previous family are still present, both the propagated terms and the notes.
    7. DONE This gene has a an annotation with a GO term (BP) that was deleted in 2010. Family ID: PTHR11139 (ATM - RELATED); Term ID: GO:0006260 (DNA Replication); Gene ID: Q13535 (UniProt ; HUMAN_ATR). GoQuick history available here: http://www.ebi.ac.uk/QuickGO/GTerm?id=GO:0006260#term=history&info=1. Appears to be fixed, unable to repeat

Thursday, July 31st

Schedule

9:00pm - 10:00pm - Daily kickoff discussion

10:00am - 12:00pm - Jamboree

12:00-1:30pm - Lunch

1:30pm - 4:30pm - Closing and wrap up

Discussion and Action Items

Participants

Name Organization

Li Ni

MGI, JAX

Suzi Lewis

LBL

Moni Munoz-Torres

LBL

Paul Thomas

USC

Judy Blake

MGI, JAX

Karen Christie

MGI, JAX

Kimberley Van Auken

WormBase, Caltech

Rama Balakrishnan

SGD, Stanford

Pascale Gaudet

SIB

Tanya Berardini

TAIR (attending remotely)

Huaiyu Mi

USC

Marc Feuermann

SIB

Remote Attendees

  • We will have Webex to share our presentations/desktop. I will send the invitation to the go-discuss mailing list.
  • For voice/audio we will use the GO Conference line (1866 953 9688, code: 801561#).


Group photo

Group Photo