Open krchristie opened 2 years ago
krchristie
commented
2 years ago For the moment, I'll leave this annotation as is as it provides the desired export if perhaps not an appropriate representation in the graph. Here's the model ID and name:
ID: gomodel:62b4ffe300000582 Name: mouse Upf3b gene regulates neural progenitor cell behaviour and neuronal outgrowth
krchristie
commented
2 years ago Is this an acceptable way to annotate? I had thought it would NOT be since the same molecule of Ufp3b cannot be in two places at the same time. [Note that in this case, the "neuronal cell body" does include the "nucleus", but if I had wanted to made an otherwise identical annotation to "neurite", then the two terms "neuronal cell body" and "neurite" are mutually exclusive.
Turning on the reasoner answered this question. It is invalid:
tmushayahama
commented
2 years ago @krchristie thanks for the feedback. We somehow forgot to remove that feature. It was still under construction and was bringing confusion. @vanaukenk do you remember what was cloneActivity use case and why we removed it.
As I remember the clone activity in this case is not cloning (which was the confusion), but pre-filling in the values with text from the copied activity so that a curator doesn't have to type again. So it is "Prefill Form Values from This Model" feature than clone Activity
krchristie
commented
2 years ago I am fine with the behavior to "Prefill Form Values", and the pre-filling values part is actually working great.
The problem is that when you edit any of those values and click "Save", it adds those changes (without deleting the previous values) to the existing model you cloned, rather than creating a new model with the modified value(s). This behavior occurs regardless of whether you edit in the pop-up window, or dismiss the pop-up window and make changes in the entry form on the left side of the page, which is also prefilled with the values of the selected annotation to clone.
I advocate strongly for having a "Clone" action that prefills values including evidence into the entry form available as it is an enormous time saver for curators. Personally, I think it is fine to call it Clone (or Duplicate, something short) and have it prefill the values into the entry form. This is similar to the behavior of the comparable workflow in Protein2GO, where there is an icon (with mouseover text of "Make another annotation like this one") to perform this action; when you click the icon, the entry form is prefilled with the values of the selected annotation.
I think it would be simpler if the values were just prefilled into the Entry form on the left hand side, without having a pop-up window at all.
The "Clone Activity" option from the menu accessed by clicking on the 3 dots on the far right side of the Form editor, in the table of activities on the right side of the Form editor, shown in screen shot below, seems to behave oddly:
ACTUAL BEHAVIOR
When you select the "Clone Activity" option, it pops open a window that you can edit, but when you click "Save", it looks in the "Form Editor" like it has changed the existing annotation WITHOUT cloning anything.
However, if you look at the "Graph Editor", you see this, where the 2nd box from the left now contains two GO CC terms:
.
Is this an acceptable way to annotate? I had thought it would NOT be since the same molecule of Ufp3b cannot be in two places at the same time. [Note that in this case, the "neuronal cell body" does include the "nucleus", but if I had wanted to made an otherwise identical annotation to "neurite", then the two terms "neuronal cell body" and "neurite" are mutually exclusive.
The Annotation Preview DOES produce what I want, which is two complete annotations that differ by the CC term, but otherwise have all the same extensions and annotations (top and bottom annotations below):
EXPECTED BEHAVIOR
The behavior that I would expect, and was actually hoping for, when I clicked the "Clone Activity" option that is at the upper level of the activity row in the Table of Annotations is to clone the ENTIRE ANNOTATION, not just one portion of it. What I am trying to do is to add an identical annotation that differs only in the CC term, but otherwise has the exact same extensions and evidence as the original.
If it is acceptable to clone just the one term, the existing behavior is fine and produces the Annotation Preview output that I want. However, I am quite suspicious that this behavior is NOT a good idea with respect to the Graph that is produced.