modify building CDS object from 10x-type data

ebi-gene-expression-group / monocle-scripts

Command line scripts wrapping functions in monocle3

8 stars 1 forks source link

modify building CDS object from 10x-type data #7

Closed a-solovyev12 closed 4 years ago

a-solovyev12 commented 4 years ago

the current version of the package (0.0.5) breaks when attempting to parse data from 10x genomics directory
this PR modifies parsing the raw data to account for different input formats
we also reduce repeated code, and account for optional presence of row/column labels in metadata files.

ktpolanski commented 4 years ago

Hey guys, I took a look at your pull request and would feel more comfortable proceeding with my existing code base. Hopefully that's okay, sorry about the lack of communication earlier. There are still some minor things in your implementation (e.g. dropping headers for metadata text imports) that would need tweaking, but I will mirror your superior dimensional compatibility logic in my version.

pinin4fjords commented 4 years ago

Great, thank you for the response. Look forward to testing the fix soon.

a-solovyev12 commented 4 years ago

@ktpolanski No problem, but could you avoid merging your PR before we run our tests on the fix you are going to make? Need to make sure the script does not break when parsing 10x-type data, which is an important use case in one of our projects.

ktpolanski commented 4 years ago

All the merging power is in nh3's hands anyway, but feel free to do any testing you may need. I've added a 10x-style mtx+tsv combo to the tests, and also create dummy names for gene annotations.

pinin4fjords commented 4 years ago

Closing in favour of https://github.com/ebi-gene-expression-group/monocle-scripts/pull/8