See https://github.com/bcgsc/utrtargets/blob/master/reference/wget.sh for the
generation of the reference files.

As for the files in the `metadata` folder, they were collected from 3 sources:

1. The MANIFEST file from the CGHub. As of this writing, the CGhub is no longer
   available.

2. The Google BigQuery table at
   https://bigquery.cloud.google.com/table/isb-cgc:tcga_seq_metadata.GCS_listing_27apr2016,
   which included the exact locations of RNA-Seq data on the GCS. As of this
   writing, this table has been updated to
   https://bigquery.cloud.google.com/table/isb-cgc:tcga_seq_metadata.GCS_listing_24jun2016,
   and the 27apr2016 version is no longer available.

3. The Google BigQuery table at
   https://bigquery.cloud.google.com/table/isb-cgc:tcga_seq_metadata.RNAseq_FastQC,
   which included the read length information extracted by FastQC for each
   sample.