cc: @mbrush

A refactoring of the annotation file format together with a refactoring of the Java code we are using to work with the annotations. This will allow us to exploit onset data better for rare disease
We now have a concrete plan for the data model for common disease that will allow us to start curating and developing algorithms to exploit temporal data. I would say this is a focus for the second half of the R24. Related to monarch-initiative/dipper#454

Consider adding a feature that will help to revise annotations if we obsolete a term in the HPO.

• Show a list of diseases with the obsoleted term
• Allow the user to "update" the annotations (Note -- this should almost always be done with manual inspection, and so we should allow something like a todo list. It should be done with manual inspection, because if we are obsoleting a term, then probably there was something suboptimal about the annotations as well!)
• The tool should create a list of all diseases that have an annotation to the obsolete term.
• The curator should then be able to open up the files for each of the diseases, reannotate them as usual, and then close the issue.

I am trying to add additional frequency info to OMIM:610443 for this ticket (obophenotype/human-phenotype-ontology#4163). There is already many of these terms listed from other sources but I am updating with a new PMID with frequency info.
Tried adding microcephaly (which is already annotated to the disease, but I am adding a second time for frequency info).
1st issue: I added the term manually using the standard gui feature. (I.e., I filled out Microcephaly, PCS, Frequency=5%, Source=PMID:18628315 . Everything added to the entry EXCEPT frequency.
Once the new entry had been added, I tried to update the frequency manually. I have done it twice and no frequency info has been added. Upon reviewing the list, it appears that '5%' frequency had been added to 'Short stature' and not 'microcephaly'. @kingmanzhang @pnrobinson

The licence in the pom states its Apache2 whereas the LICENCE file states BSD 3-clause.

Currently this just goes to the console

 } catch (IOException e) {
            logger.error(String.format("Unable to parse HPO OBO file at %s", hpoPath.getAbsolutePath() ));
            logger.error(e,e);
                throw new PhenoteFxException(String.format("Unable to parse HPO OBO file at %s [%s]", hpoPath.getAbsolutePath(),e.toString()));
        }
        if (on

Use this API
https://monarchinitiative.org/annotate/text

If an entry is already open, then open-M (by MIM number) does not work)

It would be nice to show a small symbol whether there are unsaved changes.

Add right-clickable context -menu
After right click, three choices appear

1. is present (changes the content of the NOT column to an empty string, which is the default, and means that the indicated feature is present in the disease)
2. NOT (changes the content of the NOT column to "NOT", which means that the indicated feature is ABSENT in the disease)
3. Cancel -- close the menu with no action.

If the user clicks is-present or NOT but this value was already there, then no action needs to be taken.

This is the "NOT" column @FXML
private TableColumn<PhenoRow, String> negationCol;

in PhenotePresenter

• Update pointer to current file
• Cause a small dialog to appear with the disease and disease database ID before adding first phenotype

Terms such as Chronic (Clinical course except for Onset) need to go into the Modifier column -- add some better way of doing this.

I do not see the point letting user annotating multiple disease at once. Shall we enforce that only one disease file is opened and to be annotated? @pnrobinson

Instead, open a warning message about the error and suggest to the user what they should do.

Leigh and I talked about some layout changes that can help curation:

1. get rid of Disease ID and Disease Name columns. Instead, show it as the title of the table. Also create a dropdown menu to select diseases for curation.

2. The HPO tree is very useful, it should always be there. So shrink the table to the left and add the HPO tree on the right.

3. Add last_edited by.

@pnrobinson comments?

In some cases, our annotations use out of date labels for the disease name.
For instance, OMIM:209950 was previously called Atypical mycobacteriosis, familial.
Now it is called IMMUNODEFICIENCY 27A
https://omim.org/entry/209950
It would be nice to be able to automatically update all of the disease names to some new label, for instance using a command from the Edit menu.

@ielis

HI, I was testing the v,0.4.4 and used the text mining tool. The fetching of terms worked(!), but there is an issue. There is no where to add the PMID ID on the text mining tool. I am aware of the "Source" data entry spot on the main page, but previously, I had only used that when adding a single HPO term.

So, it allows me to use the 'Fetch text mining' tool. It allows me to add text, analyze, and add selected terms. If you hit 'Confirm & close", you get the error, "Warning-attempting to update annotation without valid PMID". However, at that point, there is no way to add a PMID and it deletes all of your work. This is not ideal.

Could you either restrict opening the fetch text mining tool until after you have added a 'source' or have another way to add a source after the fact? Much data will be lost otherwise.

If you add text to the 'Text-mining analysis' window and hit 'Analyze!', the cancel button doesn't do anything. The only way to get out that is to hit 'Add selected terms'. If you don't want to add anything, you just don't select anything.

Either remove the 'cancel' button or make it actually cancel.

It would be nice to have HPOWorkbench start the Makefile in the HPO src/ontology subdirectory. This would mean that it would be easy to create a new term in Protege, update the hp.obo file, and add the annotation to an annotation record without having to run the makefile from the command line etc. This feature would only work for people with write rights to the hp.obo, but would save a lot of time.

I can reproduce it in entry OMIM:146255
I click on a certain frequency entry to update it but I get a different entry in the dialog.

PTableView class is never used.

the widget opens but when I try to send data to the server nothing happens.

similar to HPO Workbench

Unfortunately, this got broken by the new functions, which make it impossible to mark the entire row (I think). Probably we need to replace this by marking the first field in a row, or by putting some symbol in from of the row.

The paths to external resources are harrd-coded in the downloaders. These are better kept in an external resource file. Also, what’s up with all the empty configuration.properties files in the various packages?

to fix security issue

Hi @pnrobinson-
I downloaded the latest release of PhenoteFX, but it won't open any of the files.
These are my settings.

Could you help troubleshoot?

When I try to open from MIM ID or just open a file, nothing loads.

Thanks!

This is the same bug that affected the frequency column--probably the same fix will work.

Exception in thread "JavaFX Application Thread" java.lang.NullPointerException
at org.monarchinitiative.phenotefx.gui.editrow.EditRowFactory.showPublicationEditDialog(EditRowFactory.java:45)
at org.monarchinitiative.phenotefx.gui.main.PhenotePresenter.lambda$null$40(PhenotePresenter.java:1016)

Consider the best UI element to do this (menu vs. popup a dialog that will show the hierarchy something like in the HPO workbench but just for the modifer terms).

I checked the file by hand and did not find the second id (617830) anywhere.

The current method for inputting the old phenote files is way too reliant on a function in PhenotePresenter called populateTable. Transfer this functionality to a separate class that can be tested and that will provide sensible error messages if an input file has the wrong format.

https://github.com/monarch-initiative/hpo-annotation-data/issues/256

Started a new annotation that did not exist, but I used the digit code instead of OMIM:#######. Could you either restrict what is entered or have a list of what is supposed to be used?

Thanks

When the user adds a Freqeuncy Ontology term, e.g.,
https://hpo.jax.org/app/browse/term/HP:0040281,
then the label (Very frequent) gets added instead of the HPO term id (HP:0040281).
This is not the correct format and causes errors in downstream pipelines with phenol.

The only reason to add a term that is already present is if there is new metadata. Therefore, show whether a term is already there to prevent unnecessary duplication.

We should add autocomplete for clinical course (e.g., episodic) to the Modifier field.

This field should be cleared if the user closes an entry and opens a new entry.

hphenote.HPenote should that be simply ‘hphenote’?

Add some flag to the menu that will switch to the common disease mode.

This eveidence code is no longer in use.

see https://github.com/monarch-initiative/hpo-annotation-data/issues/223
Add some window or widget that will allow us to look up these numbers by entering a percentage.

exception when cancel choosing folder for annotation data

In PhenotePresenter class the setOnEditCommit() methods requiring an EventHandler the EventHandler code can be replaced with a lambda. E.g.

phenotypeIDcol.setOnEditCommit(
new EventHandler<TableColumn.CellEditEvent<PhenoRow, String>>() {
@OverRide
public void handle(TableColumn.CellEditEvent<PhenoRow, String> event) {
String hpoid = event.getNewValue();
if (HPOValidator.isValid(hpoid)) {
((PhenoRow) event.getTableView().getItems().get(event.getTablePosition().getRow())).setPhenotypeID(event.getNewValue());
}
event.getTableView().refresh();
}
}
);

Can be replaced with:

phenotypeIDcol.setOnEditCommit(
event -> {
String hpoid = event.getNewValue();
if (HPOValidator.isValid(hpoid)) {
((PhenoRow) event.getTableView().getItems().get(event.getTablePosition().getRow())).setPhenotypeID(event.getNewValue());
}
event.getTableView().refresh();
}
);

Do some Q/C and do not save unless Q/C tests pass--I notice that PhenoteFX let me save an annotation without a createdBy entry for instance.