Cptac

Fixed minor bugs in algorithm for filtering multiple mutations when merging omics and metadata dataframes with mutations dataframe. Update get_freq_mut function in cptac.algorithms to handle silent mutations in the renal dataset.

In this release, we added the cptac.algorithms sub-module, and the new renal CCRCC dataset. We also replaced the "sync" function with an improved "download" function, and eliminated unnecessary output during dataset download and loading. Finally, we made some minor changes to the API:
- Dataframes no longer have a .name attribute. Use descriptive variable names instead.
- Standardized all CNA/CNV names to CNV. This fixed a bug in accessing the Ovarian copy number dataframe.
- Renamed merge/join functions:
- compare_omics............................join_omics_to_omics
- append_mutations_to_omics.........join_omics_to_mutations
- append_metadata_to_omics..........join_metadata_to_omics
- New join functions:
- join_metadata_to_mutations
- join_metadata_to_metadata
- In join_omics_to_mutations, change "mutation_genes" parameter to "mutations_genes" (plural "mutations")
- In join_omics_to_omics, change "omics_df1_name" and "omics_df2_name" to "df1_name" and "df2_name"

For details on how these changes may affect usage, consult the use cases under the "docs" folder.

This release implements remote data storage, and includes the Colon and Ovarian datasets, as well as the Endometrial dataset. The package now just includes the source code, and the data files for each dataset are downloaded later from another location the first time the user loads each dataset. When a user loads a dataset, the package will also check that any data files already downloaded are up-to-date with the most current data files on the remote storage. If there are any discrepancies, the package gives the user the option of updating the files.

This release includes the full endometrial dataset, with all the utilities functions for working with it. We'll include the ovarian and colon cancer datasets in a future release, once we've set up remote data storage.

We have changed the dataset to be a class that the user instantiates, rather than a submodule that they load. As a result, the syntax for loading the dataset has changed slightly. Instead of running "import cptac.endometrial as en", the user would run two commands: first "import cptac", then "en = cptac.Endometrial()". However, manipulating the dataset thereafter uses the same syntax as before, working with the variable the dataset was assigned to, e.g. "clinical = en.get_clinical" and so on.

We have changed the syntax for the three merging functions: compare_omics, append_metadata_to_omics, and append_mutations_to_omics. Instead of separately loading the dataframes you want to merge, and then passing them to the function, you just pass a string to the function containing the name of the dataframe you want to merge, e.g. "appended = en.append_metadata_to_omics(metadata_df_name="derived_molecular", omics_df_name="phosphoproteomics")". Note that the parameter names now have "_name" added to the end. These functions no longer accept dataframes; you must pass dataframe names instead.