druggability

Requirements

• Software environment with python v3.7.6 or higher and pandas v1.0.5 or higher
• Alteration databases
• Variant file (single, or multi-sample, MAF or fusion results file; VCF not yet supported) aligned to GRCh38
• (Optional) Clinical trials metadata file(s)

Installation

Local installation

1. A software environment such as conda or miniconda is highly recommended.

2. Get the repositories:

   git clone --recurse-submodules https://github.com/ding-lab/druggability.git

   Running this command creates the main druggability directory and the subdirectory druggability_databases with the database files. (It is possible to clone the repos separately; just be sure to update DRUGDBDIR in myglobal.py.)

3. If relevant, activate the conda/miniconda environment and install additional packages: pip install -r requirements.txt.

Docker environment

1. Obtain the Dockerfile by direct download or from a "git clone" of the druggability repo and then build the image, e.g.,

   docker build -t druggability .

Running

$ ./druggability.py -h
usage: druggability.py [-h] [-o OUTPUT_FILE] [-l LOG_FILE] [-d] [-nn NORMAL_NAME]
                       [-tn TUMOR_NAME] [-fn FUSION_SAMPLE_NAME] [-at ANNOTATE_TRIALS]
                       [-ato TRIALS_AUXILIARY_OUTPUT_FILE] [--dump_trials_only]
                       [--maf VARIANT_MAF_FILE] [--basicmaf VARIANT_BASICMAF_FILE]
                       [--fusion VARIANT_FUSION_FILE]

optional arguments:
  -h, --help                         show this help message and exit
  -o OUTPUT_FILE                     alteration database matches
  -l LOG_FILE                        logfile name
  -nn NORMAL_NAME                    MAF normal sample name
  -tn TUMOR_NAME                     MAF tumor sample name
  -fn FUSION_SAMPLE_NAME             fusion sample name
  -at ANNOTATE_TRIALS                report clinical trials for this disease keyword
  -ato TRIALS_AUXILIARY_OUTPUT_FILE  clinical trials auxiliary output filename
  --maf VARIANT_MAF_FILE             variant file in maf format (requirements: tumor and normal names)
  --basicmaf VARIANT_BASICMAF_FILE   variant file in basic maf format (requirements: tumor and normal names)
  --fusion VARIANT_FUSION_FILE       variant file for fusions (requirement: tumor name)
  --dump_trials_only
  -d, --debug

Notes:

• Releases v1.2+:
  • Maf and fusion files can now be specified at the same time. The sample(s) responsible for the database matches are shown in the output tables.
  • Each input file type has been given its own command-line flag.
  • The sample names passed to the tool (-nn, -tn, -fn) need to match those appearing in the inputs. This feature can be helpful for comparing results across aliquots.
• For fusion inputs, the "fusion sample name" can be any sample, normal or tumor.
• See the examples directory for driver scripts to use as templates, particularly for analyzing a cohort of sample sets.
• The variant call files that were used in development are not able to be distributed externally.

MAF examples:

1. MAFs in ding-lab or PanCan union maf formats are automatically detected:

   $ ./druggability.py  --maf  LUAD.Somatic.050919.mnp.annot.maf -nn C3N-00169_N -tn C3N-00169_T  -l  druggability.log  -o alterations.out
   

   $ ./druggability.py  --maf  LUAD.Somatic.050919.mnp.annot.maf -nn C3N-00169_N -tn C3N-00169_T  -l druggability.log  -o alterations.out  -at chol  -ato  trials.aux
   

2. Ad hoc MAFs. Variant call files, such as those redistributed with publications, may have a different format than those listed above. In that case, druggability can read a tab-delimited file whose first 13 fields are the following:

   Hugo_Symbol	NCBI_Build	Chromosome	Start_Position	End_Position	Variant_Classification	Variant_Type	Reference_Allele	Tumor_Seq_Allele1	Tumor_Sample_Barcode	Matched_Norm_Sample_Barcode	HGVSc	HGVSp_Short
   

   and analyzed with the basicmaf option:

   $ ./druggability.py  --basicmaf  cohort.maf -nn 111 -tn 111  -l druggability.log  -o alterations.out  -at chol  -ato  trials.aux
   

   Notes:

   • Coordinates are 1-based
   • Variant_Type should be SNP, DNP, MNP, INS, or DEL

Fusion example:

• Input file must be in ding-lab format.

  $ ./druggability.py  --fusion  CPTAC_fusions_v0.1.csv.tsv -fn C3L-00165_T  -l druggability.log  -o alterations.out
  

Joint MAF/Fusion example:

```
$ ./druggability.py --basicmaf mc3.v0.2.8.PUBLIC.code.filtered.Missense.maf.CHOL --fusion TCGA_MC3.allfusion.compile.txt.CHOL.reformatted.tsv -nn TCGA-3X-AAVA-10A-01D-A41A-09 -tn TCGA-3X-AAVA-01A-11D-A417-09 -fn TCGA-3X-AAVA-01A-11R-A41I-07 -l TCGA_MC3_chol.TCGA-3X-AAVA.log -o TCGA_MC3_chol.TCGA-3X-AAVA.out -at chol -ato TCGA_MC3_chol.TCGA-3X-AAVA.aux
```