Assign degree of belief according to the evidence code that support the Go annotation

Question

[animesh_10kuet]

I want to assign some degree of belief that how the Go annotation to a particular term is supported which is represented by evidence code. There are different type of evidence codes:

1. Experimental Evidence codes (IDA,IEP,IGI,IMI, IPI)
2. Author Statement evidence codes (NAS,TAS)
3. Curatorial Statement codes (IC,ND)
4. Computational Analysis evidence codes (ISS, RCA)

How do i give the degree of belief value to one evidence code compare to others? For example: what will be the value of belief for IDA compare to IC and what will be the value of degree of value for IC ? and so on.

How can i assign all the degree of belief values for all the evidence codes? Is there any standard way for it?

[Jean-Karim Heriche]

Different people have different ideas on the matter so this is usually arbitrary. In practice, you assign a numerical score that reflects your confidence in the annotation type being adequate support for the annotation. In general, the goal is to modulate the influence of annotations from specific evidences so the scores can be considered as weights. How you integrate these scores into your data analysis pipeline depends on the framework you're using.

[User 59]

Cross posted to StackOverflow: http://stackoverflow.com/questions/38167948/assign-degree-of-belief-according-to-the-evidence-code-that-support-the-go-annot