decouple() was defined to allow access to all statistics in one place. For now, it supports the following statistics, all designed under the same evaluation standards:
run_gsva(); wrapper design.
run_mean() scratch design.
run_pscira() scratch design.
run_scira(); scratch design.
run_viper(); wrapper design.
All of the above functions are referred to as the decoupleR statistics family in the documentation, therefore they provide documentation that allows you to always see all the available options when asking for help from any member. For instance, ?decouple will show in the See also section all the statistics mentioned above.
convert_to_ variants family was implemented to allow centralized documentation of part of the preprocessing necessary to run a given statistic.
Since all statistics require at least defining a matrix (e.g expression matrix) and a network (e.g regulation network), it is proposed to use a centralized documentation that is inherited to all statistics through the roxygen2 tag @inheritParams.
@inheritParams .decoupler_mat_format or @inheritParams mat_format will allow you to simply inherit the description of the matrix parameter
@inheritParams .decoupler_network_format or @inheritParams network_format will allow you to inherit the description of the network parameter as well as the definition of parameters that serve as metadata columns of a network, for example, the definition of the column .source, .target, .mor or .likelihood.
It should be noted that all the metadata columns in the network must include the prefix . to easily identify them among the rest of the parameters.
Continuous integration
The use of Github Actions was incorporated to carry out the following tasks:
Evaluate the compatibility of the package in the different operating systems (necessary requirement to submit the package to Bioconductor); linux, windows and macOs.
Some additional checks are included to verify that the package meets Bioconductor's requirements, such as a style guide, biocViews, among others.
Calculate code coverage.
Run unit-testing.
Deploy web-page.
Documents
In order to be an open source package, the following commonly used documents were included:
Main code changes
decouple()was defined to allow access to all statistics in one place. For now, it supports the following statistics, all designed under the same evaluation standards:run_gsva(); wrapper design.run_mean()scratch design.run_pscira()scratch design.run_scira(); scratch design.run_viper(); wrapper design.decoupleR statisticsfamily in the documentation, therefore they provide documentation that allows you to always see all the available options when asking for help from any member. For instance,?decouplewill show in theSee alsosection all the statistics mentioned above.convert_to_ variantsfamily was implemented to allow centralized documentation of part of the preprocessing necessary to run a given statistic.matrix(e.g expression matrix) and anetwork(e.g regulation network), it is proposed to use a centralized documentation that is inherited to all statistics through the roxygen2 tag@inheritParams.@inheritParams .decoupler_mat_formator@inheritParams mat_formatwill allow you to simply inherit the description of thematrixparameter@inheritParams .decoupler_network_formator@inheritParams network_formatwill allow you to inherit the description of thenetworkparameter as well as the definition of parameters that serve as metadata columns of a network, for example, the definition of the column.source,.target,.moror.likelihood.It should be noted that all the metadata columns in the network must include the prefix
.to easily identify them among the rest of the parameters.Continuous integration
The use of
Github Actionswas incorporated to carry out the following tasks:Documents
In order to be an open source package, the following commonly used documents were included: