# for emacs: -*- mode: sh; -*-

# This file describes how the browser for Strongyloides ratti version is built

##############################################################################
# download sequence, create UCSC sequence (DONE - 2015-07-08 - Hiram)
    mkdir -p /hive/data/genomes/strRat2/genbank
    cd /hive/data/genomes/strRat2/genbank
    rsync -L -a -P \
rsync://ftp.ncbi.nlm.nih.gov/genomes/genbank/invertebrate/Strongyloides_ratti/all_assembly_versions/GCA_001040885.1_S_ratti_ED321/ ./

    faSize GCA_001040885.1_S_ratti_ED321_genomic.fna.gz
# 43150242 bases (254073 N's 42896169 real 20204330 upper 22691839 lower)
#    in 135 sequences in 1 files
# Total size: mean 319631.4 sd 1818227.7 min 181 (LN609527.1)
#    max 16759152 (LN609529.1) median 3409
# %52.59 masked total, %52.90 masked real

    mkdir /hive/data/genomes/strRat2/ucsc
    cd /hive/data/genomes/strRat2/ucsc

    # this is a structured assembly:
    ~/kent/src/hg/makeDb/doc/worms/ucscCompositeAgp.pl ../genbank/GCA_*assembly_structure/Primary_Assembly
    ~/kent/src/hg/makeDb/doc/worms/unplaced.pl ../genbank/GCA_*assembly_structure/Primary_Assembly
    ~/kent/src/hg/makeDb/doc/worms/unlocalized.pl ../genbank/GCA_*assembly_structure/Primary_Assembly

    faSize *.fa
# 43150242 bases (254073 N's 42896169 real 42896169 upper 0 lower)
#    in 135 sequences in 5 files
# Total size: mean 319631.4 sd 1818227.7 min 181 (chrUn_LN609527v1)
#    max 16759152 (chr2) median 3409
# %0.00 masked total, %0.00 masked real

#############################################################################
#  Initial database build (DONE - 2015-07-08 - Hiram)

    cd /hive/data/genomes/strRat2
    cat << '_EOF_' > strRat2.config.ra
# Config parameters for makeGenomeDb.pl:
db strRat2
clade worm
# genomeCladePriority 68
scientificName Strongyloides ratti
commonName S_ratti_ED321
assemblyDate Sep. 2014
assemblyLabel Wellcome Trust Sanger Institute S. ratti genome project
assemblyShortLabel S. ratti ED321
orderKey 19049
mitoAcc none
fastaFiles /hive/data/genomes/strRat2/ucsc/*.fa
agpFiles /hive/data/genomes/strRat2/ucsc/*.agp
# qualFiles none
dbDbSpeciesDir worm
photoCreditURL http://www.bristol.ac.uk/biology/people/mark-e-viney/index.html
photoCreditName Mark Viney/University of Bristol
ncbiGenomeId 3496
ncbiAssemblyId 356801
ncbiAssemblyName S_ratti_ED321
ncbiBioProject 62033
genBankAccessionID GCA_000208845.1
taxId 34506
'_EOF_'
    # << happy emacs

    # verify sequence and AGP are OK:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev -fileServer=hgwdev \
         -stop=agp strRat2.config.ra) > agp.log 2>&1
    # *** All done!  (through the 'agp' step)
    # real    0m10.283s

    # then finish it off:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev \
       -fileServer=hgwdev -continue=db strRat2.config.ra) > db.log 2>&1
    # real    0m35.987s

    # check in the trackDb files created and add to trackDb/makefile

##############################################################################
# cpgIslands on UNMASKED sequence (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/cpgIslandsUnmasked
    cd /hive/data/genomes/strRat2/bed/cpgIslandsUnmasked

    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku -buildDir=`pwd` \
       -tableName=cpgIslandExtUnmasked \
          -maskedSeq=/hive/data/genomes/strRat2/strRat2.unmasked.2bit \
             -workhorse=hgwdev -smallClusterHub=ku strRat2) > do.log 2>&1
    # real    1m26.695s

    cat fb.strRat2.cpgIslandExtUnmasked.txt
    # 228 bases of 43150242 (0.001%) in intersection

#############################################################################
# cytoBandIdeo - (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/cytoBand
    cd /hive/data/genomes/strRat2/bed/cytoBand
    makeCytoBandIdeo.csh strRat2

#########################################################################
# ucscToINSDC table/track (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/ucscToINSDC
    cd /hive/data/genomes/strRat2/bed/ucscToINSDC

    ~/kent/src/hg/utils/automation/ucscToINSDC.sh \
       ../../genbank/GCA_*assembly_structure/Primary_Assembly

    awk '{printf "%s\t0\t%d\n", $1,$2}' ../../chrom.sizes \
         | sort > name.coordinate.tab
    join name.coordinate.tab ucscToINSDC.txt | tr '[ ]' '[\t]' \
         > ucscToINSDC.bed
    # verify all names are coming through, should be same line count:
    wc -l *
    #   135 name.coordinate.tab
    #   135 ucscToINSDC.bed
    #   135 ucscToINSDC.txt

    cut -f1 ucscToINSDC.bed | awk '{print length($0)}' | sort -n | tail -1
    # 22
    # use the 22 in this sed
    sed -e "s/21/22/" $HOME/kent/src/hg/lib/ucscToINSDC.sql \
         | hgLoadSqlTab strRat2 ucscToINSDC stdin ucscToINSDC.bed
    checkTableCoords strRat2
    # should cover %100 entirely:
    featureBits -countGaps strRat2 ucscToINSDC
    # 43150242 bases of 43150242 (100.000%) in intersection

#########################################################################
# fixup search rule for assembly track/gold table (DONE - 2015-07-08 - Hiram)
    hgsql -N -e "select frag from gold;" strRat2 | sort | head -3
LN609396.1
LN609397.1
LN609398.1

LN609[0-9]*(\.1)?

    hgsql -N -e "select frag from gold;" strRat2 | sort | tail -2
LN609529.1
LN609530.1

    # verify this rule will find them all or eliminate them all:
    hgsql -N -e "select frag from gold;" strRat2 | wc -l
    # 135

    hgsql -N -e "select frag from gold;" strRat2 \
       | egrep -e 'LN609[0-9]*(\.1)?' | wc -l
    # 135

    hgsql -N -e "select frag from gold;" strRat2 \
       | egrep -v -e 'LN609[0-9]*(\.1)?' | wc -l
    # 0

    # hence, add to trackDb/worm/strRat2/trackDb.ra
searchTable gold
shortCircuit 1
termRegex LN609[0-9]*(\.1)?
query select chrom,chromStart,chromEnd,frag from %s where frag like '%s%%'
searchPriority 8

##########################################################################
# running repeat masker (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/repeatMasker
    cd /hive/data/genomes/strRat2/bed/repeatMasker
    time (doRepeatMasker.pl -buildDir=`pwd` \
        -bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
        -smallClusterHub=ku strRat2) > do.log 2>&1
    # real    7m41.593s

    cat faSize.rmsk.txt
    #  43150242 bases (254073 N's 42896169 real 41011065 upper 1885104 lower)
    #    in 135 sequences in 1 files
    #  Total size: mean 319631.4 sd 1818227.7 min 181 (chrUn_LN609527v1)
    #    max 16759152 (chr2) median 3409
    #  %4.37 masked total, %4.39 masked real

    egrep -i "versi|relea" do.log
    # RepeatMasker version open-4.0.5
    #    January 31 2015 (open-4-0-5) version of RepeatMasker
    # CC   RELEASE 20140131;

    time featureBits -countGaps strRat2 rmsk
    # 1885105 bases of 43150242 (4.369%) in intersection
    # real    0m0.360s

    # why is it different than the faSize above ?
    # because rmsk masks out some N's as well as bases, the count above
    #   separates out the N's from the bases, it doesn't show lower case N's

##########################################################################
# running simple repeat (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/simpleRepeat
    cd /hive/data/genomes/strRat2/bed/simpleRepeat
    time (doSimpleRepeat.pl -buildDir=`pwd` -bigClusterHub=ku \
        -dbHost=hgwdev -workhorse=hgwdev -smallClusterHub=ku \
        strRat2) > do.log 2>&1
    # real    12m19.219s

    cat fb.simpleRepeat
    # 1174848 bases of 43150242 (2.723%) in intersection

    # using the Window Masker result as indicated below

##########################################################################
# CREATE MICROSAT TRACK (DONE - 2015-07-08 - Hiram)
     ssh hgwdev
     mkdir /cluster/data/strRat2/bed/microsat
     cd /cluster/data/strRat2/bed/microsat
     awk '($5==2 || $5==3) && $6 >= 15 && $8 == 100 && $9 == 0 {printf("%s\t%s\t%s\t%dx%s\n", $1, $2, $3, $6, $16);}' \
	../simpleRepeat/simpleRepeat.bed > microsat.bed
    hgLoadBed strRat2 microsat microsat.bed
    #	Read 99 elements of size 4 from microsat.bed

##########################################################################
## WINDOWMASKER (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/windowMasker
    cd /hive/data/genomes/strRat2/bed/windowMasker
    time (doWindowMasker.pl -buildDir=`pwd` -workhorse=hgwdev \
        -dbHost=hgwdev strRat2) > do.log 2>&1
    # real    1m56.347s

    # Masking statistics
    cat faSize.strRat2.cleanWMSdust.txt
    # 43150242 bases (254073 N's 42896169 real 19033520 upper 23862649 lower)
    #    in 135 sequences in 1 files
    # Total size: mean 319631.4 sd 1818227.7 min 181 (chrUn_LN609527v1)
    #    max 16759152 (chr2) median 3409
    # %55.30 masked total, %55.63 masked real

    cat fb.strRat2.rmsk.windowmaskerSdust.txt
    # 1749886 bases of 43150242 (4.055%) in intersection

    # using this Window Masker result for final masking::
    cd /hive/data/genomes/strRat2
    #   you can safely ignore the warning about fields >= 13
    twoBitMask bed/windowMasker/strRat2.cleanWMSdust.2bit \
       -add bed/simpleRepeat/trfMask.bed  strRat2.2bit

    # measure the final masking:
    twoBitToFa strRat2.2bit stdout | faSize stdin > faSize.strRat2.2bit.txt
    cat faSize.strRat2.2bit.txt
    # 43150242 bases (254073 N's 42896169 real 19028732 upper 23867437 lower)
    #    in 135 sequences in 1 files
    # Total size: mean 319631.4 sd 1818227.7 min 181 (chrUn_LN609527v1)
    #    max 16759152 (chr2) median 3409
    # %55.31 masked total, %55.64 masked real

    # and reset the symlink
    rm /gbdb/strRat2/strRat2.2bit
    ln -s /hive/data/genomes/strRat2/strRat2.2bit /gbdb/strRat2/strRat2.2bit

##########################################################################
# cpgIslands - (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/cpgIslands
    cd /hive/data/genomes/strRat2/bed/cpgIslands
    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku \
      -workhorse=hgwdev -smallClusterHub=ku strRat2) > do.log 2>&1 &
    # real    1m27.387s

    cat fb.strRat2.cpgIslandExt.txt
    # 228 bases of 43150242 (0.001%) in intersection

#########################################################################
# augustus - (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/augustus
    cd /hive/data/genomes/strRat2/bed/augustus
    #  XXX this is not specifically correct, the species caenorhabditis
    #      is not necessarily accurate here
    time (doAugustus.pl -buildDir=`pwd` -bigClusterHub=ku \
      -species=caenorhabditis -dbHost=hgwdev \
         -workhorse=hgwdev strRat2) > do.log 2>&1
    # real    53m12.171s

    cat fb.strRat2.augustusGene.txt
    # 1701236 bases of 43150242 (3.943%) in intersection

#########################################################################
# genscan - (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat2/bed/genscan
    cd /hive/data/genomes/strRat2/bed/genscan
    time (doGenscan.pl -buildDir=`pwd` -workhorse=hgwdev -dbHost=hgwdev \
      -bigClusterHub=ku strRat2) > do.log 2>&1
    # real    7m44.826s

    cat fb.strRat2.genscan.txt
    #   852569 bases of 43150242 (1.976%) in intersection

    cat fb.strRat2.genscanSubopt.txt
    #   924308 bases of 43150242 (2.142%) in intersection

########################################################################
# Create kluster run files (DONE - 2015-07-08 - Hiram)

    cd /hive/data/genomes/strRat2
    # numerator is strRat2 gapless bases "real" as reported by:
    head -1 faSize.strRat2.2bit.txt
# 43150242 bases (254073 N's 42896169 real 19028732 upper 23867437 lower)
#    in 135 sequences in 1 files
    # numerator is 'real' base count
    # denominator is hg19 gapless bases as reported by:
    #   featureBits -noRandom -noHap hg19 gap
    #     234344806 bases of 2861349177 (8.190%) in intersection
    # 1024 is threshold used for human -repMatch:
    calc \( 42896169 / 2861349177 \) \* 1024
    # ( 42896169 / 2861349177 ) * 1024 = 15.351386

    # ==> use -repMatch=100 since 30 or 50 masks too much
    cd /hive/data/genomes/strRat2
    time blat strRat2.2bit \
         /dev/null /dev/null -tileSize=11 -makeOoc=jkStuff/strRat2.11.ooc \
        -repMatch=100
    # Wrote 6255 overused 11-mers to jkStuff/strRat2.11.ooc
    #  real    0m0.934s

    # there are no non-bridged gaps, don't need this
    #   check non-bridged gaps to see what the typical size is:
#     hgsql -N -e 'select * from gap where bridge="no" order by size;' strRat2 \
#       | ave -tableOut -col=7 stdin
# min Q1 median Q3 max mean N sum stddev
# 78831 79056 79362 79503 79503 79162.5 4 316650 320.991

    # note the minimum non-bridged gap size is 78,831
#    gapToLift -verbose=2 -minGap=50000 strRat2 jkStuff/strRat2.nonBridged.lft \
#         -bedFile=jkStuff/strRat2.nonBridged.bed

    # survey sizes:
    n50.pl chrom.sizes
#       reading: chrom.sizes
#       contig count: 135, total size: 43150242, one half size: 21575121
# cumulative    N50 count       contig  contig size
# 16759152        1       chr2    16759152
# 21575121 one half size
# 28452716        2       chr1    11693564

#############################################################################
# GENBANK AUTO UPDATE (TBD - 2015-06-09 - Hiram)
    ssh hgwdev
    cd $HOME/kent/src/hg/makeDb/genbank
    git pull
    # /cluster/data/genbank/data/organism.lst shows:
    # #organism               mrnaCnt estCnt  refSeqCnt
    #  Pristionchus pacificus  97      37470   0

    # edit etc/genbank.conf to add strRat2 just before priPac1

# strRat2 (P. pacificus)
strRat2.serverGenome = /hive/data/genomes/strRat2/strRat2.2bit
strRat2.clusterGenome = /hive/data/genomes/strRat2/strRat2.2bit
strRat2.ooc = /hive/data/genomes/strRat2/jkStuff/strRat2.11.ooc
strRat2.lift = /hive/data/genomes/strRat2/jkStuff/strRat2.nonBridged.lft
strRat2.refseq.mrna.native.pslCDnaFilter  = ${ordered.refseq.mrna.native.pslCDnaFilter}
strRat2.refseq.mrna.xeno.pslCDnaFilter    = ${ordered.refseq.mrna.xeno.pslCDnaFilter}
strRat2.genbank.mrna.native.pslCDnaFilter = ${ordered.genbank.mrna.native.pslCDnaFilter}
strRat2.genbank.mrna.xeno.pslCDnaFilter   = ${ordered.genbank.mrna.xeno.pslCDnaFilter}
strRat2.genbank.est.native.pslCDnaFilter  = ${ordered.genbank.est.native.pslCDnaFilter}
strRat2.refseq.mrna.native.load = yes
strRat2.refseq.mrna.xeno.load  = yes
strRat2.refseq.mrna.xeno.loadDesc = yes
# DO NOT NEED genbank.mrna.xeno except for human, mouse
strRat2.genbank.mrna.xeno.load = no
strRat2.genbank.est.native.load = yes
strRat2.genbank.est.native.loadDesc = no
strRat2.downloadDir = strRat2
strRat2.perChromTables = no

    git commit -m "Added strRat2 - P. pacificus refs #15209" etc/genbank.conf
    git push
    # update /cluster/data/genbank/etc/:
    make etc-update

    screen      #  control this business with a screen since it takes a while
    cd /cluster/data/genbank

    time ./bin/gbAlignStep -initial strRat2
    # logFile: var/build/logs/2015.07.02-11:39:01.strRat2.initalign.log
    # real    91m20.362s

    #   To re-do, rm the dir first:
    #     /cluster/data/genbank/work/initial.strRat2

    # load database when finished
    ssh hgwdev
    cd /cluster/data/genbank
    time ./bin/gbDbLoadStep -drop -initialLoad strRat2
    # logFile: var/dbload/hgwdev/logs/2015.07.06-09:51:31.strRat2.dbload.log
    # real    18m34.687s

    # enable daily alignment and update of hgwdev
    cd ~/kent/src/hg/makeDb/genbank
    git pull
    # add strRat2 to:
    #   vi etc/align.dbs etc/hgwdev.dbs
    git commit -m "Added strRat2 - Pristionchus pacificus refs #15209" \
      etc/align.dbs etc/hgwdev.dbs
    git push
    make etc-update

#########################################################################
# all.joiner update, downloads and in pushQ - (TBD - 2015-06-22 - Hiram)
    cd $HOME/kent/src/hg/makeDb/schema
    # fixup all.joiner until this is a clean output
    joinerCheck -database=strRat2 -tableCoverage all.joiner
    joinerCheck -database=strRat2 -times all.joiner
    joinerCheck -database=strRat2 -keys all.joiner

    cd /hive/data/genomes/strRat2
    time makeDownloads.pl strRat2 > downloads.log 2>&1
    # real    13m42.027s

    #   now ready for pushQ entry
    mkdir /hive/data/genomes/strRat2/pushQ
    cd /hive/data/genomes/strRat2/pushQ
    makePushQSql.pl strRat2 > strRat2.pushQ.sql 2> stderr.out
    #   check for errors in stderr.out, some are OK, e.g.:
    # WARNING: hgwdev does not have /gbdb/strRat2/wib/gc5Base.wib
    # WARNING: hgwdev does not have /gbdb/strRat2/wib/quality.wib
    # WARNING: hgwdev does not have /gbdb/strRat2/bbi/qualityBw/quality.bw
    # WARNING: strRat2 does not have seq
    # WARNING: strRat2 does not have extFile
    # WARNING: strRat2 does not have estOrientInfo
    # WARNING: strRat2 does not have mrnaOrientInfo

    #   copy it to hgwbeta
    scp -p strRat2.pushQ.sql qateam@hgwbeta:/tmp
    ssh qateam@hgwbeta "./bin/x86_64/hgsql qapushq < /tmp/strRat2.pushQ.sql"

    #   in that pushQ entry walk through each entry and see if the
    #   sizes will set properly

#############################################################################
# LIFTOVER TO strRat1 (DONE - 2015-07-08 - Hiram )
    mkdir /hive/data/genomes/strRat2/bed/blat.strRat1.2015-07-08
    cd /hive/data/genomes/strRat2/bed/blat.strRat1.2015-07-08
    # -debug run to create run dir, preview scripts...
    doSameSpeciesLiftOver.pl \
	-buildDir=`pwd` \
	-bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
  -ooc=/hive/data/genomes/strRat2/jkStuff/strRat2.11.ooc -debug strRat2 strRat1
    # Real run:
    time (doSameSpeciesLiftOver.pl -buildDir=`pwd` \
      -bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
       -ooc=/hive/data/genomes/strRat2/jkStuff/strRat2.11.ooc strRat2 strRat1) \
		> do.log 2>&1
    #	real    3m12.837s

    # verify it works on genome-test

#############################################################################
# improve common name (DONE - 2015-07-24 - Hiram)

    hgsql -e 'update dbDb set organism="Threadworm" where name="strRat2";' hgcentraltest

#############################################################################