BEL 2.0 - proposed language updates
Natalie Catlett
Dear All,
BEL v2.0 is coming this summer – we plan to have a full language specification available in a few weeks. (Parsing and compiler support for BEL v2.0 will require the next generation BEL Framework, a work in progress, see - https://github.com/OpenBEL/bel.rb/ )
Proposed updates to the BEL language are detailed on the OpenBEL wiki - http://bit.ly/1vtAPWp. Please take a look and comment on the proposed changes – in particular, how well do the proposed representations handle cases you’ve run into and are there alternate solutions that should also be considered. Specific areas of enhancement include:
· Variants - expand representation of sequence variants across DNA, RNA, and protein
· Post-translational modifications – move modification type to namespace, allowing use of external vocabularies and addition of modification types without requiring language updates
· Cellular location – representation of distinct cellular pools of an abundance and connect translocation terms to these pools
· Protein activity functions – streamline activity functions to act(), enable representation of specific activity type via namespace value
· Protein cleavage fragments – enable functional composition of cleavage fragments by amino acid sequence range
Thank you to everyone who has contributed formally or informally to the “BEL wish list” since the initial release of v1.0.
Best,
Natalie
Dexter Pratt
The proposals are very thoughtful and I like them a lot.
I'll space out my comments in several posts - In many cases, I'd like to explore whether the ideas could be further generalized without making things unmanageable.
And as I said at the BEL workshop, many of the additions are shifting BEL towards a level of precision such that the full language is not readily writable or readable. I think this is unavoidable - the biology that we need to capture is complex and its expression must reflect that irreducible complexity. So we need to think of new metaphors for interacting with the knowledge, mechanisms that allow us to look at understandable views without compromising the underlying statements of fact.
- Dexter
Dexter Pratt
- is there any reason to limit the loc operator to cellular and sub-cellular structures?
- It seems to me that the abundance of insulin in the bloodstream and its abundance in muscle tissue are valid locations for a translocation expression.
- The same could be said of cell types - such that we can express cause and effect between cell types, such as secretion of X by mast cells causes increased surface expression of Y on endothelial cells.
- That last thought is incomplete - the statement might only be appropriate for a particular tissue type or structure type
- So can we combine locations in some way? "Cell surface of lung endothelium"
- Is there any reason why activities and processes shouldn't also have location modifiers?
- One case to consider is a transcription factor that must be both modified AND translocated in order to be active
On Monday, May 19, 2014 3:05:50 PM UTC-4, Natalie Catlett wrote:
Natalie Catlett
Hi Dexter,
I don’t see any reason that the loc() modifier should be limited to cellular/subcellular structures. It seems reasonable to consider tissues/cell types as locations. However, we need think through the best practices for when these locations should be specified in the BEL term with loc(), versus as a statement context annotation.
As proposed, I don’t think we can combine locations to form more complex location descriptions. This is an interesting idea and should perhaps get evaluated for a future round of language updates.
I think it makes sense to limit loc() to abundances; it is an expansion/refinement of translocations, which only apply to abundances. Are there use cases where location should be applied to biological processes/pathologies? In the case of the transcription factor – or a kinase that has different targets in the cytoplasm vs. the nucleus, the inner abundance term for an activity can have a loc() modifier: act(p(X, loc(nucleus))).
Natalie
Natalie Catlett
On Monday, May 19, 2014 3:05:50 PM UTC-4, Natalie Catlett wrote: