This is a major bug for the mpnstpdx dataset and when fixed, may result in better predictions.

The current implementation of mpnstpdx/03_get_drug_response_data.R, assumes that the "new_data" files are exclusively single drug experiments. This assumption causes additional rows that are combo data to be integrated into single drug data (those with equivalent compound name are all grouped together). This assumption also leads to missing drug combination data in the pdx drug combinations file as the excel files are never extracted to this location.

Synapse ID Referenced: syn25831974

The highlighted columns are the only ones extracted in the "new_data" portion of the code.
The highlighted rows represent different drugs/combinations. The first and third rows are mistakenly grouped together in the current code.
)