vadr.config

# Format of this config file:
# - one line per 'options key': per specific set of v-annotate.pl options to
#   use for a set of models
# - each line has 3 fields: 'options_key' 'model_dir' 'options'
# - first two fields are white space delimited, (that is, first token is
#   'options key' (no spaces allowed) and second token is 'model_dir' (no spaces allowed)
#   All remaining text is combined to make field 3 (field 3 will contains whitespace)
#
dengue    $VADRINSTALLDIR/vadr-models-flavi    --split --cpu 1 --group Dengue --nomisc --noprotid --mkey flavi -r
hcv       $VADRINSTALLDIR/vadr-models-flavi    --split --cpu 1 -r --mkey flavi --group HCV
flavi     $VADRINSTALLDIR/vadr-models-flavi    --split --cpu 1 -r --nomisc
norovirus $VADRINSTALLDIR/vadr-models-calici   --split --cpu 1 --group Norovirus --nomisc --noprotid --mkey calici -r
calici    $VADRINSTALLDIR/vadr-models-calici   --split --cpu 1 -r --nomisc 
sarscov2  $VADRINSTALLDIR/vadr-models-sarscov2 --split --cpu 4 -s -r --nomisc --lowsim5seq 6 --lowsim3seq 6 --alt_fail lowscore,insertnn,deletinn --glsearch
corona    $VADRINSTALLDIR/vadr-models-corona   --split --cpu 1 -s -r --nomisc --lowsim5seq 6 --lowsim3seq 6 --alt_fail lowscore,insertnn,deletinn --glsearch 
flu       $VADRINSTALLDIR/vadr-models-flu      --split --cpu 2 -r --atgonly --alt_fail extrant5,extrant3 --xnocomp --nomisc --forcegene
rsv       $VADRINSTALLDIR/vadr-models-rsv      --split --cpu 1 -r --xnocomp --nomisc 
mpxv      $VADRINSTALLDIR/vadr-models-mpxv     --split --cpu 1 --glsearch --minimap2 -s -r --nomisc --r_lowsimok --r_lowsimxd 100 --r_lowsimxl 2000 --alt_pass discontn,dupregin --s_overhang 150
#
# Rule for how we determine <s> value for --mkey <s> to pass to v-annotate.pl
# inside v-scan.pl:
# 1. It is '--mkey <s>' if '--mkey <s>' exists in the 'options' string (field 3)
# 2. Else it is 'options_key' value
# 
# This means that multiple, different 'options_key' values can use the
# same model directory.  This allows us to have a different line for
# 'options_key' values 'norovirus' and 'calici'.  for example, but
# have them both use --mkey calici for the --cls_only classification
# stage inside v-scan.pl. Any sequences that match best to 'norovirus'
# will be rerun using v-annotate.pl with the options string for the
# line starting with 'norovirus'. Any sequences matching to
# non-norovirus models in the 'calici' library will be rerun using
# v-annotate.pl with the options string for the line starting with
# 'calici'.
# 
# For sequences to 'match best' to norovirus, they must
# match to a model in the 'calici' library that has either a name,
# group or subgroup value that is 'norovirus' (*after removing special
# characters and converting to lowercase*, so a name, group or
# subgroup values of 'NOROvirus' 'norovirus!' or 'Noro-virus' would
# all 'match best' to norovirus). The v-scan.pl program checks
# that at least one model in each library $okey (e.g. 'calici') meets
# this criteria for every $okey2 (e.g. 'norovirus') that uses $okey as
# its model key.
#
# About the --split --cpu <n> options:
# * --split is used to parallelize processing in v-annotate.pl
#   if you want to run on only 1 CPU, remove `--split --cpu <n>`
#   from each line
# * --cpu <n> controls how many threads will run, you may want
#   to change this depending on how much available RAM you have:
# * suggested amount of RAM per CPU for each virus:
#   sarscov2: 2Gb 
#   flu: 4Gb
#   dengue, hcv, flavi, norovirus, calici, corona, mpxv: 8Gb
#   rsv: 16Gb