Sorry, this PR is the first step of filtering where the filter_ki branch builds on the filter_qc_ki5 branch. PR #40 is trying to merge filter_ki2 into filter_ki. I submitted those PRs out of order since this one will be PR #41.
This PR makes a lot of documentation changes, explaining parameters and how to use the gimap_filter() function.
It also builds the groundwork for the filters working together.
All supported/possible filters are first set to be NULL and then based on user input for which filter(s) to run, these variables are overwritten/remain NULL.
A list of all possible filters is built, and then we use the reduce() function to cbind these filters together. This approach ignores NULLs and returns a df with column(s) of TRUEs and FALSEs.
Then the min_n_filters parameter is used together with rowSums to find out which pgRNA constructs are flagged by at least that minimum number of filters that would result in the pgRNA constructs being removed from the dataset -- creating a consensus or combined filter.
Sorry, this PR is the first step of filtering where the
filter_kibranch builds on thefilter_qc_ki5branch. PR #40 is trying to mergefilter_ki2intofilter_ki. I submitted those PRs out of order since this one will be PR #41.This PR makes a lot of documentation changes, explaining parameters and how to use the
gimap_filter()function.It also builds the groundwork for the filters working together.
reduce()function tocbindthese filters together. This approach ignores NULLs and returns a df with column(s) of TRUEs and FALSEs.min_n_filtersparameter is used together withrowSumsto find out which pgRNA constructs are flagged by at least that minimum number of filters that would result in the pgRNA constructs being removed from the dataset -- creating a consensus or combined filter.