New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

SEP 026 -- Add a link from Location to Sequence #59

Open
udp opened this Issue Jun 28, 2018 · 11 comments

Comments

Projects
None yet
4 participants
@udp
Member

udp commented Jun 28, 2018

@udp udp added this to the SBOL 2.3 milestone Jun 28, 2018

@cjmyers cjmyers changed the title from SEP 027 -- Add a link from Location to Sequence to SEP 026 -- Add a link from Location to Sequence Jun 28, 2018

@palchicz

This comment has been minimized.

Contributor

palchicz commented Jul 16, 2018

Some more background on this SEP would be helpful. Can the SEP include examples of when a ComponentDefinitions would have multiple Sequences? It's unclear if this is referring to the case where the sequences are of different versus the same encoding. If of the same encoding, is this meant for when sequence annotations are being used to describe features. What about the case of discontinuous regions?

@cjmyers

This comment has been minimized.

Contributor

cjmyers commented Jul 16, 2018

I let @udp update the SEP, but I can add a bit more context here. In a discussion at HARMONY, we identified some cases where multiple sequences are appropriate, such as:

  • Different encodings of the same thing (e.g. smiles and inchi)
  • DNA sequence + methylation patterns

The encoding may be different, but it is not clear which Sequence a SequenceAnnotation refers to. This is the reason this reference is needed. In the case of discontinuous regions, I would assume that each Location making up the region would refer to the same Sequence. However, this is not required. So, you could also have an annotation that annotates two sequences simultaneously to indicate where the same feature shows up in each sequence.

@udp

This comment has been minimized.

Member

udp commented Jul 16, 2018

Can the SEP include examples of when a ComponentDefinitions would have multiple Sequences?

SBOL already allows multiple sequences per ComponentDefinition and has since SBOL 2.0. I don't think SEPs need to justify existing parts of the data model.

@cjmyers

This comment has been minimized.

Contributor

cjmyers commented Jul 16, 2018

@palchicz

This comment has been minimized.

Contributor

palchicz commented Jul 17, 2018

@cjmyers That explanation makes sense. Just to repeat what you are saying, the use case this is trying to get at is when there are multiple encodings.

@udp Maybe to clarify my question a bit more, the standard leaves open the possibility for multiple sequences with the same encoding (from section 7.7).

If a ComponentDefinition refers to more than one Sequence with the same encoding, then the elements of these Sequence objects SHOULD have equal lengths.

I guess that's the scenario I want to understand more, and if this proposed change applies.

@cjmyers

This comment has been minimized.

Contributor

cjmyers commented Jul 17, 2018

Ah, that's a good question. I think that the best practice would be that if no sequence reference is provided that this means the SequenceAnnotations apply to all sequences at the same specified location. If only one Sequence is being annotated, then one should provide a sequence reference.

@palchicz

This comment has been minimized.

Contributor

palchicz commented Jul 22, 2018

Awesome! Thanks @cjmyers for the explanation.

This SEP doesn't sound controversial to me. I can send out an email to the DEV list reminding people to take a look after the SEP is amended with some of the background/motivation material we've just discussed. If no one dissents, I'll move it to a vote.

@cjmyers

This comment has been minimized.

Contributor

cjmyers commented Jul 22, 2018

Sounds good. @udp please update as per this thread.

@palchicz

This comment has been minimized.

Contributor

palchicz commented Sep 11, 2018

@cjmyers and @udp one final question. Would the new sequence property be required or optional. A reading of the SEP makes it sound like it would be required, but I wanted to double check.

@cjmyers

This comment has been minimized.

Contributor

cjmyers commented Sep 11, 2018

It must be optional otherwise this would need to be SBOL 3. We do not allow SBOL 2.x changes to invalidate earlier versions of SBOL 2.y where y <= x.

@palchicz palchicz added Accepted and removed Draft labels Oct 11, 2018

@NeilWipat

This comment has been minimized.

Collaborator

NeilWipat commented Oct 11, 2018

Update as of COMBINE 2018

Has been accepted, James McLaughlin to edit into spec

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment