Annotations to Catalytic activity with IPI
Annotations to Catalytic activity and its child terms using IPI
The IPI (Inferred from Physical Interaction) evidence code, is used where an annotation can be supported from interaction evidence between the gene product of interest and another molecule (see evidence code documentation: http://www.geneontology.org/GO.evidence.shtml#ipi). While the IPI evidence code is frequently used to support annotations to terms that are children of GO:0005488; binding, it is thought unlikely by the Binding WG that enough information can be obtained from a binding interaction to support an annotation to a term that is a chid of GO:0003824; catalytic activity. Such IPI annotations to child terms of GO:0003824 may need to be revisited and corrected.
Many groups have already agreed that such annotations should not be valid. SGD has 88 such annotations and we wanted to examine them before agreeing to make this a hard rule.
Here are our findings:
- Many of the annotations to catalytic activity branch were incorrect use of IPI evidence code. These have been or will be corrected. We agree that IPI alone cannot support annotation to a catalytic activity term and we can agree to make that a rule.
- However, there are several cases where a combination of evidence supports annotation to a catalytic activity term, at least as currently defined, sometimes with the "contributes_to" qualifier. Although the evidence codes are listed together in SGD displays, that is not the case in the GAF and that causes confusion. GO has never made a rule that each piece of evidence has to support an annotation on its own. In these cases, we feel that IPI in combination with other evidence codes is the accurate representation of the evidence that led researchers to make these conclusions and that those annotations should be kept. In addition, sometimes multiple lines of evidence come from different papers and we should allow that as well.
There are some two subunit complexes where one subunit is catalytic but functions in a complex of at least two subunits where a second subunit is responsible for substrate targeting. Without the specificity subunit, the catalytic activity is absent or severely impaired. All of the annotations in question are for the substrate specificity subunits. As currently defined, annotations of "contributes_to" the following catalytic activity terms by IPI is one of the contributing lines of evidence that led researchers to make the conclusion that each of the genes listed below is involved in the catalytic activity. Other lines of evidence included IMP, IGI, and/or ISS. No single line of evidence is individually sufficient; it is the combination that leads the researchers to make the conclusion that the annotation represents.
- ubiquitin-protein ligase activity (RAD18 and UBR1) - endopeptidase activity (SOM1) - protein serine/threonine phosphatase activity (RTS1)
So while we agree that IPI cannot be the sole evidence to support an annotation to a catalytic term, we feel that annotations IPI can be used with other lines of evidence to represent the literature accurately.
For substrate specificity subunits in general, possibly there is an ontology change that would better represent what is going on. We'll use the E2/E3 ubiquitin ligases as the example. There are two completely different type of E3's (see attached figure).
In one type (E3 HECT) there is a single gene product that both recognizes (binds) the substrate and catalyzes the reaction. In the second type (E2/E3 RING complex), the catalysis and substrate binding activities are present in separate gene products. It might be a more accurate representation of what is going on to create new terms to represent the two distinct parts that are involved in this reaction, the catalytic transfer and the substrate recognition/binding, where the term representing the complete activity would have has_part relationships to each of these two new terms.
- small conjugating protein ligase activity (GO:0019787) -- (is_a) ubiquitin-protein ligase activity (GO:0004842) --- (has_part) ubiquitin-protein ligase ubiquitin transfer to substrate (GO:new1) --- (has_part) ubiquitin-protein ligase substrate binding (GO:new2)
If this activity can be represented this way, then HECT type E3's could be annotated to term (GO:0004842) representing the entire activity, comprising both substrate binding and catalytic transfer. For the E2/E2 type, the E2 subunit would be annotated to the GO:new1 term while the E3 RING subunit would be annotated to the GO:new2 term.