Documentation needed

[FerriolCalvet]

• Input from deepUMIcaller

  • Which VCF and BAM files to take
  • Which sample/batch of samples to take

• Data sources section

  • Download Ensembl VEP cache
  • Where to download CADD scores from
  • Oncodrive3D datasets
  • Signatures
  • dNdScv covariates and reference regions

• How to interpret the output? (include it in the tools section)

  • Sanity checks
  • omega vs omegagloballoc
  • Site selection values
  • Signatures

• How to run?

  • Scenario 1: first run to check general mutation statistics
  • Scenario 2: full run to look at all the results
  • Scenario 3: run to look only at signatures
  • Scenario 4: run with a custom set of mutations

• Examples section

  • Blacklist mutations
  • Custom annotation
  • Gene grouping
  • "omega hotspots" regions definition

• Documenting sources

  • Citing tools in the Citations.md, not only final analysis results but also intermediate processing tools.
  • Pointing to the repos/papers of all the used analysis tools.
  • Documenting different containers used in the process.
    • Ensure that the containers we use for each process are fully reproducible from the Dockerfile that is available in the lab repo of containers. Right now they are not.

• Define LICENSE

[FerriolCalvet]

Explain all the possible details to turn on and off.

or another option is to do a tutorial of which different types of runs the user could do

[FerriolCalvet]

From Miguel's comment:

for future versions, it could be nice to have a brief descripition of each output, e.g. in a table format:

Output Directory/File	Description  
sumannotation	        Aggregated mutation annotations?  
germline_somatic	Annotation of germline vs somatic variants? 

[FerriolCalvet]

• PMEAN and PSTD related
  If available in the input VCF, deepCSA parses for each mutation the mean position in the read and the standard deviation of this position. If not available, columns PMEAN and PSTD are set to -1.

[FerriolCalvet]

UPDATE FEATURE GROUPS definition

the only change is in the way of inputting the information for the groups of samples and so on
it has changed from this:

features_table_dict      =  ['"unique_identifier" : "SAMPLE_ID"',
                              '"groups_of_interest" : [ ["BLADDER_LOCATION"], ["BLADDER_LOCATION", "SEX"], ["BLADDER_LOCATION", "SEX", "SMOKING_STATUS"] ]' join(',\t').trim()

to these two variables:

features_unique_identifier               = "SAMPLE_ID"
features_groups_list                     = '[ ["BLADDER_LOCATION"], ["BLADDER_LOCATION", "SEX"], ["BLADDER_LOCATION", "SEX", "SMOKING_STATUS"] ]'

Done!

[FerriolCalvet]

• Document QCs
  • mutation density
  • omega
  • mutational profile