belbio / bep Goto Github PK

Time

How does the second week of November sound? That should give us time to scramble to write and pre-review any remaining BEPs 😄

Agenda

Definitely on the docket:

• BEP-0004 - equivalence relationship (#11)
• BEP-0005 - regular expression namespaces (#12)
• BEP-0006 - gene modification syntax (#13)
• BEP-0007 - numeric annotations (#16)

Still needs BEP:

• Intake/endocytosis function (#7)
• Style Guide (#17)

• Choose which BEPs should be included
• Choose release date
• Gain experience and write a document about release procedure

@cebel and @sgebel would like an updated version of the openbel.org site - right now it only shows the BEL 1.0 and 2.0 specifications. We need to make an update to incorporate all 2.1 changes

The BEL v2.1 release corresponds to this GitHub project: https://github.com/belbio/bep/projects/1. It includes the following BEPs:

• BEP-0001 - populationAbundance function (#2)
• BEP-0003 - noCorrelation relationship (#5)
• BEP-0004 - equivalence relationship (#11)
• BEP-0005 - regular expression namespaces (#12)

To Do

• Create new YAML file on https://github.com/belbio/bel_specifications/tree/master/specifications
• Write post on Medium

I'm not absolutely sure if this is a good idea, but defining namespaces by the regular expressions defined by Identifiers.org would be a good way to defer semantic validation for the most tricky namespaces (such as dbSNP)

While BEL.bio may become the reference implementation for the BEL language, it would be prudent to encourage the language and community towards agnosticism in regards to these sorts of technical details

I've already started writing an opinionated style guide here: https://github.com/pharmacome/curation/blob/master/style-guide.rst

While some of it may be a bit picky, we should have some minimum community standards for BEL scripts. Example: is there a preferred maximum line length? Is it written anywhere that spacing is important?

@ncatlett I think we would best write this together

I'd like to see numeric/quantitative BEL annotations specifically supported as numbers. Use cases would include filtering knowledge (e.g., edges with 'cohort size' > 20, 'p-value' < 0.01) and potentially weighting edges based on significance or effect size.

In BEL Script 1.0, namespaces and annotations could be defined as a url (pointing to a belns or belanno file), a list of values, or a pattern/regular expression (see https://wiki.openbel.org/display/BLD/Control+Records). To some extent, this representation need is covered by patterns, but the original specification does not specifically cover handing of numeric annotations as numbers.

I am also interested in specifying time annotations, but am not sure if they belong in a separate BEP since units (e.g., minutes, hours, days, years) also need to be considered.

Earlier this week, @johnbachman mentioned he wanted to do this. Now, we're working on curating how Tau antibodies are interacting with Tau, and we have the same need. Let's use this issue thread to start collating examples and discussion. CC: @AldisiRana

One idea:

p(HGNC:YFG, domain(start_end))

Also it would be good to show some examples with their equivalences to InterPro terms as well

Much like secretion() and cellSurfaceExpression() it would be nice to have an intake() function that's syntactic sugar around the translocation() function that represents the opposite direction of movement from secretion().

This would have some nice examples for viruses pulling themselves into cells, for endocytosis of chemicals like neurotransmitters, and for chemicals activating receptors which then get endocytosis-ed as well

BEL documents have either been written in 1.0, 2.0, a mixture of 1.0/2.0, or 2.0.0+

Should BEL documents specify what version they're using? This wouldn't affect 1.0 or 2.0, but would make differentiating future versions with minor/major updates much more easily.

Perhaps with something in the document section of the BEL script like SET BEL = "1.0.0".

Another question: how would we deal with documents that are a mixture of BEL 1.0/BEL 2.0?

1. We have a specific need to add the modification type Acylation/Acyl to the default namespace. (this general term seems like a good fit to describe acylated Ghrelin, as this is how it is frequently referred to in the literature. (The specific modification appears to be o-octanoylation (MOD_00295) or o-decanoylation (MOD_00390), but many/most papers do not refer to the modification in this way)

2. We need a mechanism to make additions and other changes to the BEL default namespace. This could be a BEP, but the level of discussion/review and the review and update cycle timing needs may be different.

See also related issue belbio/bel_resources#98

• binding of antibody to a cell?
• something with parasites?

Once this is done, we can accept BEP-0001.

It might be the case that text evidence is not available. Sometimes, it comes from a database or a table that has no text. This would also need some thorough guidelines as well.

Receiving a web page that says:

404 Not Found
Code: NoSuchKey
Message: The specified key does not exist.
Key: bep
RequestId: 32A5B3D108FA8081
HostId: G9Po1vh9BUO0s4Tkw7k4Lb5WsgDmzI26jAm4dEP6AeTSzv/UbQxWzvCHI+nvYPxVPuYRct4RRBY=

Like the noCorrelation and equivalentTo relationships, a partOf relationship would allow much better expression of the semantics common to other knowledge assemblies. See FamPlex as a good example. Sometimes, partOf has the inverse, but broader semantics as the hasComponent relationship.

We should consider the semantics of partOf and hasPart as described in the Gene Ontology as well:

• http://www.geneontology.org/page/ontology-relations#partof
• http://www.geneontology.org/page/ontology-relations#haspart

In BEL 1.0/2.0 this made sense:

complex(FPLX:9_1_1) hasComponent p(HGNC:HUS1)
complex(FPLX:9_1_1) hasComponent p(HGNC:RAD1)
complex(FPLX:9_1_1) hasComponent p(HGNC:RAD9A)

It might be better to write in the other direction:

p(HGNC:HUS1)  partOf complex(FPLX:9_1_1)
p(HGNC:RAD1)  partOf complex(FPLX:9_1_1)
p(HGNC:RAD9A) partOf complex(FPLX:9_1_1) 

Reasoning over part of

Basically, any direct paths containing only partOf and isA imply partOf.

A partOf B and B isA C implies A partOf C

For proteins, this would describe a hierarchy of complexes.

p(X) partOf complex(Y)
complex(Y) isA complex(Z)

# Infer:
p(X) partOf complex(Z)

• Example
• Logically consistent

A partOf B and B partOf C implies A partOf C

For proteins, this would describe a complex formed of other complexes.

p(X) partOf complex(Y)
complex(Y) partOf complex(Z)

# Infer:
p(X) partOf complex(Z)

• Example
• Logically consistent

Does FamPlex have an example for either of these first two headers (@johnbachman)? I will write a script to check the https://github.com/sorgerlab/famplex/blob/master/relations.csv to see.

A isA B and B partOf C implies A partOf C

p(HGNC:PRKAA1) isA p(FPLX:AMPK_alpha)
p(HGNC:PRKAA2) isA p(FPLX:AMPK_alpha)
p(FPLX:AMPK_alpha) partOf p(FPLX:AMPK)

# Infer:
p(HGNC:PRKAA1) partOf p(FPLX:AMPK)
p(HGNC:PRKAA2) partOf p(FPLX:AMPK)

• Example
• Logically consistent

Does this make sense? Should we infer all AMPK complexes have PRKAA1? I don't think that's what FamPlex is trying to say. If p(X1) isA p(X) and p(X2) isA p(X) and p(X) partOf complex(Y) then should we should infer that either X1 or X2 are present in Y?