addRun.sh

#!/bin/sh
#
# this scripts sets up a run in the database   
# to use it - e.g.
# addRun.sh 150630_D00390_0232_AC6K0WANXX 
# jumps through various hoops due to idiosyncracies of 
# \copy and copy
# only intended for importing GBS runs at this stage
#
function get_opts() {

help_text="\n
      usage : addRun.sh -r run_name\n
      example (dry run) : ./addRun.sh -n -r 151016_D00390_0236_AC6JURANXX\n
      example           : ./addRun.sh -r 151016_D00390_0236_AC6JURANXX\n
      example           : ./addRun.sh -r 171026_M02412_0043_000000000-D2N2Ua -m miseq\n
      example           : ./addRun.sh -r 190919_M02412_0124_000000000-CKPD4 -m miseq -s SQ2853\n
"

DRY_RUN=no
RUN_NAME=""
GBS_SAMPLE_LIB=""
MACHINE=hiseq
FORCE=no
FLOWCELL_MONIKER=""

while getopts ":nhr:m:d:s:fF:l:" opt; do
  case $opt in
    n)
      DRY_RUN=yes
      ;;
    i)
      INTERACTIVE=yes
      ;;
    r)
      RUN_NAME=$OPTARG
      ;;
    F)
      FLOWCELL_MONIKER=$OPTARG
      ;;
    l)
      LANE=$OPTARG
      ;;
    d)
      RUN_PATH=$OPTARG
      ;;
    f)
      FORCE=yes
      ;;
    m)
      MACHINE=$OPTARG
      ;;
    s)
      GBS_SAMPLE_LIB=$OPTARG
      ;;
    h)
      echo -e $help_text
      exit 0
      ;;
    \?)
      echo "Invalid option: -$OPTARG" >&2
      exit 1
      ;;
    :)
      echo "Option -$OPTARG requires an argument." >&2
      exit 1
      ;;
  esac
done

KEY_DIR=/dataset/hiseq/active/key-files
}

function check_opts() {
if [ -z "$GBS_PRISM_BIN" ]; then
   echo "GBS_PRISM_BIN not set - quitting"
   exit 1
fi

if [ -z "$RUN_NAME" ]; then
   echo -e $help_text
   exit 1
fi

if [ ! -d $RUN_PATH ]; then
   echo $RUN_PATH not found
   exit 1
fi

#if [ $MACHINE == "miseq" ]; then
#   echo "error , database import of miseq runs not currently supported"
#   exit 1
#fi

if [[ ( ! -f $RUN_PATH/SampleSheet.csv ) && ( $MACHINE == "hiseq" ) ]]; then
   echo $RUN_PATH/SampleSheet.csv not found , try using -d option to provide path to sample sheet
   exit 1
fi

in_db=`$GBS_PRISM_BIN/is_run_in_database.sh $RUN_NAME`
if [ $in_db != "0" ]; then
   if [ "$FORCE" != "yes" ]; then 
      echo "$RUN_NAME has already been added - quitting (use -f to override, e.g. adding samples)"
      exit 1 
   fi
fi

# machine must be miseq , hiseq or novaseq
if [[ ( $MACHINE != "hiseq" ) && ( $MACHINE != "miseq" ) && ( $MACHINE != "novaseq" ) && ( $MACHINE != "iseq" ) ]]; then
    echo "machine must be miseq or hiseq"
    exit 1
fi

}

function echo_opts() {
    echo "importing $RUN_NAME from $RUN_PATH"
    echo "DRY_RUN=$DRY_RUN"
}

get_opts $@

check_opts

echo_opts

## from here , process the import

rm -f /tmp/${RUN_NAME}.txt
if [ -f /tmp/${RUN_NAME}.txt ]; then
   echo "rm -f /tmp/${RUN_NAME}.txt failed - quitting"
   exit 1
fi
rm -f /tmp/${RUN_NAME}.psql
if [ -f /tmp/${RUN_NAME}.psql ]; then
   echo "rm -f /tmp/${RUN_NAME}.psql failed - quitting"
   exit 1
fi


# if this is a hiseq run, parse and import the sample sheet and do all the rest
#cp $RUN_ROOT/${RUN_NAME}/SampleSheet.csv /tmp/${RUN_NAME}.txt
#awk -F, '{if(NF>5)print}' ${RUN_PATH}/SampleSheet.csv  > /tmp/${RUN_NAME}.txt
if [ $MACHINE == "hiseq" ]; then 
   set -x
   cat ${RUN_PATH}/SampleSheet.csv | $GBS_PRISM_BIN/sanitiseSampleSheet.py -r $RUN_NAME > /tmp/${RUN_NAME}.txt
   set +x

   # check we got something non-trivial
   if [ ! -s /tmp/${RUN_NAME}.txt ]; then
      echo "error parsed sample sheet /tmp/${RUN_NAME}.txt is missing or empty"
      exit 1
   fi


   echo "


   insert into bioSampleList (xreflsid, listName, listComment)
   values(:run_name, :run_name, 'AgResearch Hiseq Run');

   delete from samplesheet_temp;


   \copy samplesheet_temp from /tmp/${RUN_NAME}.txt with  CSV HEADER

  insert into hiseqSampleSheetFact (
   biosamplelist ,
   FCID ,
   Lane ,
   SampleID ,
   SampleRef ,
   SampleIndex ,
   Description ,
   Control ,
   Recipe,
   Operator ,
   SampleProject ,
   sampleplate,
   samplewell,
   downstream_processing,
   basespace_project)
  select
   obid,
   FCID ,
   Lane ,
   SampleID ,
   SampleRef ,
   SampleIndex ,
   Description ,
   Control ,
   Recipe,
   Operator ,
   SampleProject, 
   sampleplate,
   samplewell ,
   downstream_processing,
   basespace_project
  from 
   bioSampleList as s join samplesheet_temp as t
   on s.listName = :run_name and 
   t. sampleid is not null;

  insert into biosampleob(xreflsid, samplename, sampledescription, sampletype)
   select distinct
   SampleID,
   SampleID,
   description,
   case when t.downstream_processing = 'GBS' then 'Illumina GBS Library'
   else 'Illumina Library'
   end
  from
   samplesheet_temp t where
   sampleid is not null and 
   not exists (select obid from biosampleob where samplename = t.SampleID);

  insert into biosamplelistmembershiplink (biosamplelist, biosampleob)
   select 
    l.obid,
    s.obid
   from 
    (biosampleob as s join samplesheet_temp as t on 
    t.sampleid = s.samplename ) join biosamplelist as l on 
    l.listname = :run_name  
   except
    select
    biosamplelist, 
    biosampleob
   from
    biosamplelistmembershiplink as m join biosamplelist as l on
    m.biosamplelist = l.obid and
    l.listname = :run_name ;
" > /tmp/${RUN_NAME}.psql
elif [[ ( $MACHINE == "miseq" ) || ( $MACHINE == "novaseq" ) || ( $MACHINE == "iseq" ) ]]; then
   # we don't bother with the sample sheet - just add any libraries from command line option
   echo "
   insert into bioSampleList (xreflsid, listName, listComment)
   select 
      :run_name,:run_name,'AgResearch Illumina Run'
   except
   select 
      xreflsid, listName, listComment from bioSampleList 
   where
      xreflsid = :run_name and 
      listName = :run_name and 
      listComment = 'AgResearch Illumina Run';
   " > /tmp/${RUN_NAME}.psql

   # if a sample lib provided on command line, add setup of that as well
   if [ ! -z "$GBS_SAMPLE_LIB" ]; then

      if [ -z "$FLOWCELL_MONIKER" ]; then
         echo "please supply a flowcell moniker (-F option)"
         exit 1
      fi
      if [ -z "$LANE" ]; then
         echo "please supply a lane arg (-l option)"
         exit 1
      fi

      echo "
  insert into hiseqSampleSheetFact (
   biosamplelist ,
   fcid ,
   Lane ,
   SampleID ,
   downstream_processing)
  select
   obid,
   :fcid ,
   :lane ,
   :gbs_sample_lib,
   'GBS'
  from
   bioSampleList where listName = :run_name ;

  insert into biosampleob(xreflsid, samplename, sampletype)
   select :gbs_sample_lib, :gbs_sample_lib, 'Illumina GBS Library'
   except
   select 
      xreflsid, samplename, sampletype
   from
      biosampleob
   where
      xreflsid = :gbs_sample_lib and samplename = :gbs_sample_lib and sampletype = 'Illumina GBS Library';

  insert into biosamplelistmembershiplink (biosamplelist, biosampleob)
   select
    l.obid,
    s.obid
   from
    biosampleob as s join biosamplelist as l on
    l.listname = :run_name and s.xreflsid = :gbs_sample_lib
   except
    select
    biosamplelist,
    biosampleob
   from
    biosamplelistmembershiplink as m join biosamplelist as l on
    m.biosamplelist = l.obid and
    l.listname = :run_name ;
   " >> /tmp/${RUN_NAME}.psql 
   fi # adding a mised sample library
fi # miseq 

if [ $DRY_RUN == "no" ]; then
   psql -U agrbrdf -d agrbrdf -h postgres -v run_name=\'${RUN_NAME}\' -v fcid=\'${FLOWCELL_MONIKER}\' -v lane=$LANE -v gbs_sample_lib=\'${GBS_SAMPLE_LIB}\' -f /tmp/${RUN_NAME}.psql
else
   echo " will run 
   psql -U agrbrdf -d agrbrdf -h postgres -v run_name=\'${RUN_NAME}\' -v fcid=\'${FLOWCELL_MONIKER}\' -v lane=$LANE -v gbs_sample_lib=\'${GBS_SAMPLE_LIB}\' -f /tmp/${RUN_NAME}.psql"
fi


echo "done (URL to access run is http://agbrdf.agresearch.co.nz/cgi-bin/fetch.py?obid=${RUN_NAME}&context=default )"