Skip to content

Conversation

@edawson
Copy link
Owner

@edawson edawson commented Feb 6, 2016

Enough said, really. I've been in the bad habit of developing on master.

Eric Dawson and others added 30 commits November 16, 2015 20:53
Simplification was removing deletions at the beginning of mappings. This caused
problems in merged alignments produced by banding, which generate multiple
mappings to the same node.
Due to banding, multiple passes may be required to include paths. This should
be parameterized, but for now I'm just testing.
These can handle multiple mappings per path.
Whoops, there was a bug with the last commit.
Hard to describe aside from "fixing bugs related
to normalization of complex paths over graphs".

I'm a bit closer to supporting paths in unchop.
This can take a mapping and split it into two mappings at a specified
cut point, then store these in the paths.

The rank that they are assigned is each the same as the original
element, so only via normalization are the ranks going to be made
correct.
Including cyclic paths in graphs
edawson and others added 28 commits December 20, 2015 21:39
This ensures that normalization works as expected. These requirements for
sibling set membership were heavy-handed and appear unnecessary in any of the
tests I have available (stock test and HLA genes).
Add .pre-build make dependencies and deconstructor skeleton classes.
… as otherwise it would take forever to push out nanopore reads.
…o that it may be sourced from anywhere. Also moved alignment output back to a 1000 alignment buffer
edawson added a commit that referenced this pull request Feb 6, 2016
Merge master into dev branch.
@edawson edawson merged commit ffb295f into dev Feb 6, 2016
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants