Optimizing a subset of antigens individually #85
Conversation
|
(Just seen a batch of |
|
Random thought - a simpler approach in this type of situation might be to give antigens in the subset a (much) lower weight. |
|
Yeah, this is a nice addition, although I hesitate to add stuff like this to the core Racmacs package since as you say there are so many variants of this type of thing, and anything we add to Racmacs requires maintaining and ideally proper tests etc. Plus it feels maybe quite similar to the "frozen merge" https://acorg.github.io/Racmacs/articles/merging-maps.html#frozen-merge. I had at one point started a RacmacsUtils package for stuff like this, maybe that would be a good home for it? |
|
Yeah tricky... in one way it would be cleaner to keep things like this in The last step is doing the same as a frozen merge (I think). I tried to call |
… as the merged map. Also applies the plotspec.
…h columns corresponding to stress per titer when nd values are excluded and when they are included
…ith disconnected points
…s provided, names will be taken from the row and column names when creating a map.
…es have different lengths or are missing for some points
Update to v1.1.31
Update version number
… for all antigens / sera
… for all antigens / sera
* Verbosity argument also suppresses optimization progress messages * Issue a warning if duplicate antigen or serum names are found when creating a map
Update to v1.1.33
…reviated`, `srNamesFull` removed
Update to 1.1.34
Add antigen and sera attributes: `agLineage()`, `srLineage()`, `agReassortant()`, `srReassortant()`, `agStrings()`, `srStrings()`, `agContinent()`, `agNucleotideSequences()`, `srNucleotideSequences()`
Add antigen and sera attributes: `agLineage()`, `srLineage()`, `agReassortant()`, `srReassortant()`, `agStrings()`, `srStrings()`, `agContinent()`, `agNucleotideSequences()`, `srNucleotideSequences()`
Adding nucleotide sequence test & fix typo
Add additional antigen and sera attributes
Add documentation
Add documentation (#7)
… as the merged map. Also applies the plotspec.
…cs into optimize-subset-individually
|
Was looking at reviewing this and under files changed it has a bunch of changes and e.g. in DESCRIPTION it has stuff like 1.1.30 -> 1.1.35, but in the current version of master DESCRIPTION already has 1.1.35, so I'm wondering why it's listed as a change. Any idea what's going on, seems like it's comparing to some old version of master for some reason? |
|
I rebased this branch with master, to incorporate all the recent changes. Under 'Files changed' select 'Changes from' and select the top two commits (fe5a1cb and 9428a86) to see just the changes I made. (Sorry for the extra confusion. I had thought that in rebasing I wouldn't end up with all the extra individual commits from master...) |
|
(Lavanya had needed some of the new functionality (altering point transparency) as well as the optimize subset individually function.) |
Thought this might be useful as a library function.
The function removes a subset of antigens and optimizes (yielding the 'reference' map). Then adds each antigen in the subset one by one, optimizes, and aligns to the reference. Position of each antigen in the subset is then added to the reference map.
Useful, for example, for maps containing 10s-100s of mutants.
There could be variations:
Not sure if if put this in the best file. Also, would be interested if you'd suggest doing anything differently @shwilks.