Has any implementation of GFA implemented the containment C record? about gfa-spec HOT 14 CLOSED

I will take a look at the PR and let you know if I would stick to GFA1 or migrate to GFA2.

from gfa-spec.

A kind of: https://github.com/medvedevgroup/TwoPaCo

The code is not "released" yet, though.

from gfa-spec.

Okay. Good to know. Perhaps the C record should be left as is then. If no one had implemented it, I was thinking we could possibly get rid of it. Would you like to comment on #33?

from gfa-spec.

I will take a look.

from gfa-spec.

I notice that, in commit c881263, recently merged in the specification, there is a description of C lines, which I find confusing on one point:
"This line can be used to describe mapping between segments in the graph and input sequences."

I.e. until now I thought that the C (as well as L) "from" and "to" fields were supposed to be segment names from S lines in the graph. In the note about "pan-genome" motivation for C, it sounds like one can have a C between a S and a chromosome sequence which is external, not contained in the graph.

I think it would be better to clarify this in the specification text.

from gfa-spec.

@ggonella you are right. The point is to keep the mapping between pan-genome graph and input sequences, see #7 .

from gfa-spec.

I agree with Giorgio @ggonnella that the containment C record is a relationship between two segment S records both in the GFA file. GFA doesn't currently have any concept of external IDs. I suggest changing the text like so:

 # `C` Containment line

-This line can be used to describe mapping between segments in the graph and input sequences. It is useful for graphs built from assembled genomes, e.g. pan-genome analysis graphs.
+A containment line represents an overlap between two segments where one is contained in the other.

I've created PR #41 with this change.

from gfa-spec.

@ggonnella @IlyaMinkin Please vote/comment on #41 so that I can merge it. Three 👍 are required to merge.

from gfa-spec.

@IlyaMinkin If TwoPaCo is the only implementation of GFA1 containment C records, would you consider migrating to GFA 2 so that we can deprecate the containment record of GFA 1?

from gfa-spec.

@sjackman does GFA2 specification exist somewhere? And what is the reason for deprecation of containment?

from gfa-spec.

GFA 2.0 is coming very close to a consensus. The PR is #48. Your comments are welcome.

If you are actively using GFA 1.0 containment C records in released code, then that's fine, we won't deprecate them. If however they were never implemented in released code, then I'd prefer to deprecate them.

from gfa-spec.

I do not understand, why it should be better to deprecate the C lines. They will not be there anymore in GFA2, but also the L lines will not: should we deprecate L lines too? That wouldn't make sense. C lines express things that cannot be expressed in other ways in GFA1.

from gfa-spec.

If C records had never been implemented by anyone, it would be better to encourage people to migrate to GFA 2.

from gfa-spec.

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.

from gfa-spec.