Closed CuzickA closed 2 days ago
Curation link: https://canto.phi-base.org/curs/1f5e9ca2be427dab
I would need new PHIPO terms like modulation of host DNA synthesis (promote and/or reduce), modulation of the phosphorylation done by a host kinase of a host protein (increase/reduce), gal/tumor formation I also need new experimental condition terms: γ32ATP, heterologous species saccharomyces cerevisae
[help #222199] [uuw] UniProtKB/TrEMBL A0A0D1C8C8 entry update request
[help #222200] [uuw] UniProtKB/TrEMBL B4FQV3 entry update request
[help #222201] [uuw] UniProtKB/TrEMBL Q9M6Q9 entry update request
Hi @smvelasquez, I was just having a quick look at this session and noticed that the protein-protein interactions were under the 'gene-for-gene phenotype' section. In this case they should be moved so they are under the 'pathogen-host interaction phenotype'.
I can see how this could be confusing. The curated sessions that you trained on for 'physical interactions' were based on experiments where there was a known gene for gene interaction and so the protein-protein interaction phenotype annotations were added to the gene for gene phenotype section. This will not be the case for all physical interactions.
Gene-for-gene interaction results from the genetic relationship between a pathogen and a host. More info can be found at http://prgdb.crg.eu/wiki/Gene-for-gene_interaction#:~:text=Gene%2Dfor%2Dgene%20interaction%20results,a%20pathogen%20and%20a%20host. We developed the Gene for Gene Annotation Extensions so that we could capture extra information on the compatibility of the interaction, and whether there was a recognizable pathogen effector present and/or functional host resistance gene present.
In this paper, we are curating a physical interaction between a pathogen effector (see1) and its first host target (SGT1) and so this would fall under the 'pathogen-host interaction phenotype' section. There is no mention of a genetic gene for gene relationship between different pathogen and host strains like we see in the atr1/RPP1 example #19. If there is a gene for gene relationship the authors will normally mention this within the manuscript text.
@smvelasquez when you have finished curating this paper, please submit it and then I can review it. Thanks.
@CuzickA I need to fix these things you pointed out and then I will submit it
@CuzickA I've submitted the curation :)
Checking this session
I have changed pathogen genotype from
to
as the strain information is already captured in the strain slot.
And
to
I added '+' which indicates WT
What is the best way to curate 'plant tumors' which comprise of both host and pathogen material. Currently curated with annotations related to pathogen asexual spores, may also need annotation for host morphology
@ValWood , do you think it would be better to make a new term for ‘tumor’ comprising of pathogen and host material or annotate pathogn spores and host morphology concurrently?
Deleting following annotations. These control complementation annotations are not required, as it is a given that the authors have correctly performed these controls.
I'm not sure about this. Intuitively, I would say if these "sexual spores within host structures" are tumours, I would make the term labels be about presence or absence of tumours. The definition could contain the information about how they are formed.
I think it would be better to use 'gall' than tumour.
Tumour can also refer to abnormal growth without a pathogen eg cancer
NTRs: presence or absence of galls (These are plant specific, do we want to specify that detail in the definition?)
If we move to curating with the new 'host gall' terms it looks as if some spore information could be curated from S Fig 2
The new 'host gall' term annotations would look like this
@CuzickA I agree with using the term "gall". Martin told me that they have switched to that terminology in more recent publications from the Ustilago field. If we curate data from Fig S2, would there be the option to add e.g. normal morphology?
@CuzickA I agree with using the term "gall". Martin told me that they have switched to that terminology in more recent publications from the Ustilago field. If we curate data from Fig S2, would there be the option to add e.g. normal morphology?
Thanks for your comment @smvelasquez. I will go ahead and create the PHIPO terms for 'gall'. I'm not sure about your question for Fig S2 until I read and check that part of the paper curation. I'll get back to you on this question.
I have just added these annotations for S Fig 2
Hi @smvelasquez, just following up on Fig S2
Perhaps we could create a new child term 'presence of viable pathogen sexual spores within host' under parent 'presence of pathogen sexual spores within host'
This could cover the two bits of key information 1) spores are present 2) spores are viable ( can germinate like the WT)
Please let me know what you think?
For the RNA expression annotations
I think we only want to use 1 time point in the conditions. In this case I will use 12 dpi as this is the time point described in SFig 3 and the main text. _I deleted these in preference of the WT RNA level annotations (AC 11_1122)
I also added the information for RNA present in the tassel (not upregulated like in the leaf).
I have 2 questions for @ValWood 1) Is there a formal rule about not entering more that one time point in the conditions? My understanding is that only one time point per annotation should be used. If this is correct we may need to check that this is clear in the Help documentation or short video tutorials that we plan to develop. 2) When annotating wild-type RNA levels do we need to make annotations in both the 'Wild-type RNA level' annotation type AND the 'Pathogen-host interaction phenotype'? The only novel data in the latter is the condition recording the timepoint. When annotating RNA levels to altered metagenotypes then I can see it makes sense to make the annotations in both sections.
Looks like there is also some expression data that could be curated for the see1 mutant inoculated onto maize. Unfortunately, I could not find the UniProt ids for the maize gene ids listed below.
Fig 3 I think these annotations should be deleted as expt in Fig 3 is only for WT metagenotype. The GO annotation below with NTRs should cover this.
Need to request new GO terms NTR: 'effector-mediated modulation of host DNA synthesis by symbiont' AND child term NTR: 'effector-mediated induction of host DNA synthesis' or similar.
_(AC 09_0922 Fig 3 phenotype annotations now deleted in favour of GO annotations. Need to request new GO terms).
Fig 4 Looks like we also need some new PHIPO terms for the pathogen induction of host DNA synthesis Suggest NTR: presence of pathogen induced host DNA synthesis NTR: decreased pathogen induced host DNA synthesis
I'm not sure quite where to place them in PHIPO
Any suggestions @ValWood?
Fig 6 (_AC 09_0922 Fig 6 annotations now deleted in preference of GO annotations) Edit to pathogen genotype here
the deleted region is the N-terminal secretion signal 1-21. The remaining construct is 22-157.
@CuzickA Hi, I quite agree with adding the viable spores term. I think it would capture nicely what happens in that figure.
Fig 6 (AC 09_09_22 Fig 6 annotations now deleted in preference of GO annotations)
Still working on these genotypes The first 2 could be reduced. The latter 3 still in green need to represent deltaSee1 and complementation with various constructs. I think these need to be made as multi-allele genotypes following https://github.com/PHI-base/curation/issues/29
updated to
Need to look at these genotypes again. How to create genotype with Promoter from Pit2 (Ppit2-see1). Do we need to add UniProt for Pit2?
Also having difficulty in deciding whether the 'mCherry, HA tag' should be in background or part of the genotype name.
At PomBase we do these with allele type "other"
We put 'mCherry, HA tag' in the background, but we would sometimes put in the forground if they are shown to have a substantial effect on the phenotype
At PomBase we do these with allele type "other"
We put 'mCherry, HA tag' in the background, but we would sometimes put in the forground if they are shown to have a substantial effect on the phenotype
Thanks. These genotypes are tricky, I have created multiallele genotypes deleted pathogen gene delta see1 transformed WT pathogen gene see1+ with mCherry, HA tag WT host
or same with deleted pathogen gene delta see1 transformed portion of pathogen gene see1 (promoter and signal peptide) with mCherry, HA tag WT host
I'm also beginning to wonder whether these phenotype annotations should even be made for Fig 6 (above) as the take home message is that pathogen see1 is localized to the host nucleus and cytoplasm which can be captured as GO annotations.
@ValWood, please let me know what you think? (AC 09_09_22 Fig 6 annotations now deleted in preference of GO annotations)
Fig 7 (following curation in https://canto.phi-base.org/curs/6ddeb3a009b5a0e5/ro/ and https://canto.phi-base.org/curs/f41d30e250b8f9b0/ro/) I would change the evidence to 'reporter gene study' for the Y2H and change the genotypes for see1 the SP was deleted 1-21
simplifying genotypes to
to
Hi @smvelasquez, after thinking about Fig 6 some more I am planning on deleting all the phenotype entries and just keeping the GO annotations.
My reasoning for this is that the key point is that pathogen see1 localizes to the host cytoplasm and nuclei. The difficult genotypes that I was trying to create really 'represent' WT see1 or control expts. Therefore, the expts represent the normal localization of this protein within the pathogen-host interaction and this can be represented with the GO cellular component annotations.
If the authors created a genotype with a partial deletion or AA substitution in see1 that altered the normal localization then this could be annotated as a PHI phenotype with WT metagenotype control.
Does this make sense?
_(AC 09_0922 Fig 6 annotations now deleted in preference of GO annotations)
Fig 8
change to
to
Need some new PHIPO terms here. NTR: host protein phosphorylation present with pathogen. NTR: decreased host protein phosphorylation with pathogen. NTR: increased host protein phosphorylation with pathogen. under parent 'pathogen modulation of host process phenotype'. Need to make it clear in def that in this case presence of pathogen enables host kinase to phosphorylate host protein (SGT1). May want to add AE 'assayed_using' so can indicate SGT1 is the host protein being phosphorylated. Still need to think of a better way to represent the OE Ppit2-see1 genotype. If Ppit2 just being used for OE then perhaps if can be removed from the genotype and just use over expression for the expression level.
@CuzickA I think what you are proposing makes sense and with the GO annotation, the localization data is covered
@CuzickA I was discussing with Martin about this Ustilago promoter. He told me that Pit2 is a promoter that turns itself on during biotrophy. His group didn't use it (they used another one) but the idea is to use a strong promoter to drive the expression of another effector whose promoter is weaker. So basically, Pit2 is used as a tool. I don't think we need the Uniprot code for it. It is "sort of" like a 35s, but for biotrophy only, obviously. Hope this helps
@CuzickA I was discussing with Martin about this Ustilago promoter. He told me that Pit2 is a promoter that turns itself on during biotrophy. His group didn't use it (they used another one) but the idea is to use a strong promoter to drive the expression of another effector whose promoter is weaker. So basically, Pit2 is used as a tool. I don't think we need the Uniprot code for it. It is "sort of" like a 35s, but for biotrophy only, obviously. Hope this helps
Thanks @smvelasquez thats helpful. I think I will remove the promoter detail from the genotypes and just use overexpression. Therefore changes to
edit genotype
I don't think PHI-Canto is set up for curating this type of interspecies complementation experiment so I'm going to delete the annotations for now, but leave a record of them here in this ticket.
Hi @smvelasquez, do you think we should also add 'has substrate' 'SGT1 (100282745)' to the top annotation?
@CuzickA I think it would be good to add the 'has_substrate'
@smvelasquez which AE option did you use for 'has_substrate' here?
When I try an add the information, it seems like the only relevant AE is 'phosphorylates' and that is 'has_input'. I wonder if something has changed in the AE here or maybe I have missed something??
Need new GO terms
@smvelasquez which AE option did you use for 'has_substrate' here?
When I try an add the information, it seems like the only relevant AE is 'phosphorylates' and that is 'has_input'. I wonder if something has changed in the AE here or maybe I have missed something??
@CuzickA Hi, I was trying to remember what I chose, but all I can think of is 'phosphorylates', because it is the only that would have made sense
Thanks @smvelasquez @jseager7 do you know if there have been any changes in the naming of the GO AE for 'has_input' and 'has_substrate'?
@CuzickA Not sure what you mean by "changes in the naming". If you mean whether the names of the relations have changed in the past, then apparently they have: in July 2022, PomBase changed has_substrate to has_input in the annotation extension configuration for the GO Molecular Function annotation type. Here's the relevant commit:
https://github.com/pombase/pombase-config/commit/fba0166c60e9126ca18d2e4c040415429f9fd4aa
And the linked issue:
Thanks @jseager7. It looks like this change hasn't automatically updated within the existing annotations in PHI-Canto. In this case I'll change it manually.
It looks like this change hasn't automatically updated within the existing annotations in PHI-Canto. In this case I'll change it manually.
I've contacted the PomBase team to ask how they updated the relation name, since they would've had to do the updating somehow.
It looks like this change hasn't automatically updated within the existing annotations in PHI-Canto. In this case I'll change it manually.
I've contacted the PomBase team to ask how they updated the relation name, since they would've had to do the updating somehow.
I have manually changed the 'has substrate' to 'has input' for the 'Phosphorylates' AE
@jseager7, has there been any progress on following up whether this can be done automatically for any existing annotations using this AE?
Checked today whether our UniProt gene name changes requested on 16 May 2022 had been made yet - still nothing.
Did anyone respond at all? Definitely re-send it shouldn't take this long, but it could have fallen between releases. @Antonialock when was the last UniProt release?
For curation by @smvelasquez