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

# This file describes how the browser for D. immitis WS245 version is built

##############################################################################
# download sequence, create UCSC sequence (DONE - 2015-06-24 - Hiram)
    mkdir -p /hive/data/genomes/dirImm1/ws245
    cd /hive/data/genomes/dirImm1/ws245

    wget --no-parent --timestamping -m -nH --cut-dirs=6 \
ftp://ftp.sanger.ac.uk/pub/wormbase/releases/WS245/species/PRJEB1797

    mkdir /hive/data/genomes/dirImm1/ucsc
    cd /hive/data/genomes/dirImm1/ucsc
    # WormBase contig names are of the pattern:
# >nDi.2.2.scaf00001 1 1085577
# >nDi.2.2.scaf00002 1 1069194
# >nDi.2.2.scaf00003 1 995087
#   ...

    # can't have the dots in the names
    zcat ../ws245/PRJEB1797/d_immitis.PRJEB1797.WS245.genomic.fa.gz \
        | sed -e 's/ 1 .*//; s/\./_/g;' | gzip -c > dirImm1.fa.gz

    hgFakeAgp dirImm1.fa.gz dirImm1.agp

    # photograph
    mkdir /hive/data/genomes/dirImm1/photo
    cd /hive/data/genomes/dirImm1/photo
    wget -O REPugh_Dirofilaria_immitis_1_04.jpg \
   "http://parasite.org.au/pugh-collection/JpegsStamped/Dirofilaria immitis 1 04.jpg"

    convert -geometry "400x300" REPugh_Dirofilaria_immitis_1_04.jpg \
      Dirofilaria_immitis.jpg

#############################################################################
#  Initial database build (DONE - 2015-07-09 - Hiram)

    cd /hive/data/genomes/dirImm1
    cat << '_EOF_' > dirImm1.config.ra
# Config parameters for makeGenomeDb.pl:
db dirImm1
clade worm
genomeCladePriority 68
scientificName Dirofilaria immitis
commonName Dog heartworm nematode
assemblyDate Sep. 2013
assemblyLabel Edinburgh University D. immitis genome project
assemblyShortLabel WS240/D. immitis v2.2
orderKey 4670
mitoAcc AJ537512.1
fastaFiles /hive/data/genomes/dirImm1/ucsc/dirImm1.fa.gz
agpFiles /hive/data/genomes/dirImm1/ucsc/dirImm1.agp
# qualFiles none
dbDbSpeciesDir worm
photoCreditURL http://parasite.org.au/blog/the-pugh-parasitology-collection/
photoCreditName R.E. Pugh parasitology collection
ncbiGenomeId 10757
ncbiAssemblyId PRJEB1797
ncbiAssemblyName WS240/D. immitis v2.2
ncbiBioProject 208635
genBankAccessionID PRJEB1797
taxId 6287
'_EOF_'
    # << happy emacs

    # verify sequence and AGP are OK:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev -fileServer=hgwdev \
         -stop=agp dirImm1.config.ra) > agp.log 2>&1
    # *** All done!  (through the 'agp' step)
    # real    0m26.615s

    # then finish it off:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev \
       -fileServer=hgwdev -continue=db dirImm1.config.ra) > db.log 2>&1
    # real    1m1.272s

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

##############################################################################
# cpgIslands on UNMASKED sequence (DONE - 2015-07-09 - Hiram)
    mkdir /hive/data/genomes/dirImm1/bed/cpgIslandsUnmasked
    cd /hive/data/genomes/dirImm1/bed/cpgIslandsUnmasked

    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku -buildDir=`pwd` \
       -tableName=cpgIslandExtUnmasked \
          -maskedSeq=/hive/data/genomes/dirImm1/dirImm1.unmasked.2bit \
             -workhorse=hgwdev -smallClusterHub=ku dirImm1) > do.log 2>&1
    # real    12m23.178s

    cat fb.dirImm1.cpgIslandExtUnmasked.txt
    # 27978 bases of 84923352 (0.033%) in intersection

#############################################################################
# cytoBandIdeo - (DONE - 2015-07-09 - Hiram)
    mkdir /hive/data/genomes/dirImm1/bed/cytoBand
    cd /hive/data/genomes/dirImm1/bed/cytoBand
    makeCytoBandIdeo.csh dirImm1

#########################################################################
# ucscToINSDC table/track (TBD - 2015-03-20 - Hiram)
    # this is done only on genbank assemblies
    mkdir /hive/data/genomes/dirImm1/bed/ucscToINSDC
    cd /hive/data/genomes/dirImm1/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 *
    # 25187 name.coordinate.tab
    # 25187 ucscToINSDC.bed
    # 25187 ucscToINSDC.txt

    cut -f1 ucscToINSDC.bed | awk '{print length($0)}' | sort -n | tail -1
    # 14
    # use the 14 in this sed
    sed -e "s/21/14/" $HOME/kent/src/hg/lib/ucscToINSDC.sql \
         | hgLoadSqlTab dirImm1 ucscToINSDC stdin ucscToINSDC.bed
    checkTableCoords dirImm1
    # should cover %100 entirely:
    featureBits -countGaps dirImm1 ucscToINSDC
    # 2053849526 bases of 2053849526 (100.000%) in intersection

#########################################################################
# fixup search rule for assembly track/gold table (DONE - 2015-07-09 - Hiram)
    hgsql -N -e "select frag from gold;" dirImm1 | sort | head -3
AJ537512.1
nDi_2_2_scaf00001_1
nDi_2_2_scaf00001_10

(nDi_2_2_scaf|AJ537512)[0-9]*([\._][0-1]*)?

[JP][Fp][a4][1_][C4][o1][n1][t7][i\.](g[0-9]*)?(\_[0-9]*)?

    hgsql -N -e "select frag from gold;" dirImm1 | sort | tail -2
nDi_2_2_scaf16060_1
nDi_2_2_scaf16061_1

    # verify this rule will find them all or eliminate them all:
    hgsql -N -e "select frag from gold;" dirImm1 | wc -l
    # 24243

    hgsql -N -e "select frag from gold;" dirImm1 \
      | egrep -e '(nDi_2_2_scaf|AJ537512)[0-9]*([\._][0-1]*)?' \
        | wc -l
    # 24243

    hgsql -N -e "select frag from gold;" dirImm1 \
      | egrep -v -e '(nDi_2_2_scaf|AJ537512)[0-9]*([\._][0-1]*)?' \
       | wc -l
    # 0

    # hence, add to trackDb/worm/dirImm1/trackDb.ra
searchTable gold
shortCircuit 1
termRegex (nDi_2_2_scaf|AJ537512)[0-9]*([\._][0-1]*)?
query select chrom,chromStart,chromEnd,frag from %s where frag like '%s%%'
searchPriority 8

##########################################################################
# running repeat masker (DONE - 2015-07-09 - Hiram)
    mkdir /hive/data/genomes/dirImm1/bed/repeatMasker
    cd /hive/data/genomes/dirImm1/bed/repeatMasker
    time (doRepeatMasker.pl -buildDir=`pwd` \
        -bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
        -smallClusterHub=ku dirImm1) > do.log 2>&1
    # real    193m42.028s

    cat faSize.rmsk.txt
    # 88323343 bases (3430609 N's 84892734 real 81279691 upper
    #    3613043 lower) in 16062 sequences in 1 files
    # Total size: mean 5498.9 sd 31910.5 min 200 (nDi_2_2_scaf16061)
    #    max 1085577 (nDi_2_2_scaf00001) median 754
    # %4.09 masked total, %4.26 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 dirImm1 rmsk
    # 7720275 bases of 172510819 (4.475%) in intersection
    #  real    0m6.103s

    # 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 - 2015-07-09 - Hiram)
    mkdir /hive/data/genomes/dirImm1/bed/simpleRepeat
    cd /hive/data/genomes/dirImm1/bed/simpleRepeat
    time (doSimpleRepeat.pl -buildDir=`pwd` -bigClusterHub=ku \
        -dbHost=hgwdev -workhorse=hgwdev -smallClusterHub=ku \
        dirImm1) > do.log 2>&1
    # real    33m27.044s

    cat fb.simpleRepeat
    # 2032093 bases of 84923352 (2.393%) in intersection

    # using the Window Masker result as indicated below

##########################################################################
# CREATE MICROSAT TRACK (DONE - 2015-07-09 - Hiram)
     ssh hgwdev
     mkdir /cluster/data/dirImm1/bed/microsat
     cd /cluster/data/dirImm1/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 dirImm1 microsat microsat.bed
    #	Read 211 elements of size 4 from microsat.bed

##########################################################################
## WINDOWMASKER (DONE - 2015-07-09 - Hiram)
    mkdir /hive/data/genomes/dirImm1/bed/windowMasker
    cd /hive/data/genomes/dirImm1/bed/windowMasker
    time (doWindowMasker.pl -buildDir=`pwd` -workhorse=hgwdev \
        -dbHost=hgwdev dirImm1) > do.log 2>&1
    # real    6m42.227s

    # Masking statistics
    cat faSize.dirImm1.cleanWMSdust.txt
    # 88323343 bases (3430609 N's 84892734 real 51697598 upper 33195136 lower)
    #    in 16062 sequences in 1 files
    # Total size: mean 5498.9 sd 31910.5 min 200 (nDi_2_2_scaf16061)
    #    max 1085577 (nDi_2_2_scaf00001) median 754
    # %37.58 masked total, %39.10 masked real

    cat fb.dirImm1.rmsk.windowmaskerSdust.txt
    # 3188537 bases of 88323343 (3.610%) in intersection

    # using this Window Masker result for final masking::
    cd /hive/data/genomes/dirImm1
    #   you can safely ignore the warning about fields >= 13
    twoBitMask bed/windowMasker/dirImm1.cleanWMSdust.2bit \
       -add bed/simpleRepeat/trfMask.bed  dirImm1.2bit

    # measure the final masking:
    twoBitToFa dirImm1.2bit stdout | faSize stdin > faSize.dirImm1.2bit.txt
    cat faSize.dirImm1.2bit.txt
    # 88323343 bases (3430609 N's 84892734 real 51686546 upper 33206188 lower)
    #   in 16062 sequences in 1 files
    # Total size: mean 5498.9 sd 31910.5 min 200 (nDi_2_2_scaf16061)
    #   max 1085577 (nDi_2_2_scaf00001) median 754
    # %37.60 masked total, %39.12 masked real

    # and reset the symlink
    rm /gbdb/dirImm1/dirImm1.2bit
    ln -s /hive/data/genomes/dirImm1/dirImm1.2bit /gbdb/dirImm1/dirImm1.2bit

##########################################################################
# cpgIslands - (DONE - 2015-07-15 - Hiram)
    mkdir /hive/data/genomes/dirImm1/bed/cpgIslands
    cd /hive/data/genomes/dirImm1/bed/cpgIslands
    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku \
      -workhorse=hgwdev -smallClusterHub=ku dirImm1) > do.log 2>&1 &
    # real    46m58.080s

    cat fb.dirImm1.cpgIslandExt.txt
    # 23281 bases of 84923352 (0.027%) in intersection

#########################################################################
# augustus - (DONE - 2015-07-15 - Hiram)
    mkdir /hive/data/genomes/dirImm1/bed/augustus
    cd /hive/data/genomes/dirImm1/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 dirImm1) > do.log 2>&1
    # real    49m9.685s

    cat fb.dirImm1.augustusGene.txt
    # 4613656 bases of 84923352 (5.433%) in intersection

#########################################################################
# genscan - (DONE - 2015-07-15 - Hiram)
    mkdir /hive/data/genomes/dirImm1/bed/genscan
    cd /hive/data/genomes/dirImm1/bed/genscan
    time (doGenscan.pl -buildDir=`pwd` -workhorse=hgwdev -dbHost=hgwdev \
      -bigClusterHub=ku dirImm1) > do.log 2>&1
    # real    15m20.629s

    cat fb.dirImm1.genscan.txt
    #   2387414 bases of 84923352 (2.811%) in intersection

    cat fb.dirImm1.genscanSubopt.txt
    #   2284563 bases of 84923352 (2.690%) in intersection

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

    cd /hive/data/genomes/dirImm1
    # numerator is dirImm1 gapless bases "real" as reported by:
    head -1 faSize.dirImm1.2bit.txt
# 94076581 bases (7759220 N's 86317361 real 53341517 upper 32975844 lower)
#    in 9780 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 \( 86317361 / 2861349177 \) \* 1024
    # ( 86317361 / 2861349177 ) * 1024 = 30.890665

    # ==> use -repMatch=100 since 30 or 50 masks too much
    cd /hive/data/genomes/dirImm1
    time blat dirImm1.2bit \
         /dev/null /dev/null -tileSize=11 -makeOoc=jkStuff/dirImm1.11.ooc \
        -repMatch=100
    # Wrote 8527 overused 11-mers to jkStuff/dirImm1.11.ooc
    #  real    0m1.846s

    # there are a few non-bridged gaps
    #   check non-bridged gaps to see what the typical size is:
    hgsql -N -e 'select * from gap where bridge="no" order by size;' dirImm1 \
       | 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 dirImm1 jkStuff/dirImm1.nonBridged.lft \
         -bedFile=jkStuff/dirImm1.nonBridged.bed

    # survey sizes:
    n50.pl chrom.sizes
#       reading: chrom.sizes
#       contig count: 9780, total size: 94076581, one half size: 47038290
# cumulative    N50 count       contig  contig size
# 46960459        61      Bmal_v3_scaffold61      194773
# 47038290 one half size
# 47151548        62      Bmal_v3_scaffold62      191089

#############################################################################
# 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 dirImm1 just before priPac1

# dirImm1 (P. pacificus)
dirImm1.serverGenome = /hive/data/genomes/dirImm1/dirImm1.2bit
dirImm1.clusterGenome = /hive/data/genomes/dirImm1/dirImm1.2bit
dirImm1.ooc = /hive/data/genomes/dirImm1/jkStuff/dirImm1.11.ooc
dirImm1.lift = /hive/data/genomes/dirImm1/jkStuff/dirImm1.nonBridged.lft
dirImm1.refseq.mrna.native.pslCDnaFilter  = ${ordered.refseq.mrna.native.pslCDnaFilter}
dirImm1.refseq.mrna.xeno.pslCDnaFilter    = ${ordered.refseq.mrna.xeno.pslCDnaFilter}
dirImm1.genbank.mrna.native.pslCDnaFilter = ${ordered.genbank.mrna.native.pslCDnaFilter}
dirImm1.genbank.mrna.xeno.pslCDnaFilter   = ${ordered.genbank.mrna.xeno.pslCDnaFilter}
dirImm1.genbank.est.native.pslCDnaFilter  = ${ordered.genbank.est.native.pslCDnaFilter}
dirImm1.refseq.mrna.native.load = yes
dirImm1.refseq.mrna.xeno.load  = yes
dirImm1.refseq.mrna.xeno.loadDesc = yes
# DO NOT NEED genbank.mrna.xeno except for human, mouse
dirImm1.genbank.mrna.xeno.load = no
dirImm1.genbank.est.native.load = yes
dirImm1.genbank.est.native.loadDesc = no
dirImm1.downloadDir = dirImm1
dirImm1.perChromTables = no

    git commit -m "Added dirImm1 - 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 dirImm1
    # logFile: var/build/logs/2015.07.02-11:39:01.dirImm1.initalign.log
    # real    91m20.362s

    #   To re-do, rm the dir first:
    #     /cluster/data/genbank/work/initial.dirImm1

    # load database when finished
    ssh hgwdev
    cd /cluster/data/genbank
    time ./bin/gbDbLoadStep -drop -initialLoad dirImm1
    # logFile: var/dbload/hgwdev/logs/2015.07.06-09:51:31.dirImm1.dbload.log
    # real    18m34.687s

    # enable daily alignment and update of hgwdev
    cd ~/kent/src/hg/makeDb/genbank
    git pull
    # add dirImm1 to:
    #   vi etc/align.dbs etc/hgwdev.dbs
    git commit -m "Added dirImm1 - 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=dirImm1 -tableCoverage all.joiner
    joinerCheck -database=dirImm1 -times all.joiner
    joinerCheck -database=dirImm1 -keys all.joiner

    cd /hive/data/genomes/dirImm1
    time makeDownloads.pl dirImm1 > downloads.log 2>&1
    # real    13m42.027s

    #   now ready for pushQ entry
    mkdir /hive/data/genomes/dirImm1/pushQ
    cd /hive/data/genomes/dirImm1/pushQ
    makePushQSql.pl dirImm1 > dirImm1.pushQ.sql 2> stderr.out
    #   check for errors in stderr.out, some are OK, e.g.:
    # WARNING: hgwdev does not have /gbdb/dirImm1/wib/gc5Base.wib
    # WARNING: hgwdev does not have /gbdb/dirImm1/wib/quality.wib
    # WARNING: hgwdev does not have /gbdb/dirImm1/bbi/qualityBw/quality.bw
    # WARNING: dirImm1 does not have seq
    # WARNING: dirImm1 does not have extFile
    # WARNING: dirImm1 does not have estOrientInfo
    # WARNING: dirImm1 does not have mrnaOrientInfo

    #   copy it to hgwbeta
    scp -p dirImm1.pushQ.sql qateam@hgwbeta:/tmp
    ssh qateam@hgwbeta "./bin/x86_64/hgsql qapushq < /tmp/dirImm1.pushQ.sql"

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

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

    hgsql -e 'update dbDb set organism="Dog heartworm" where name="dirImm1";' hgcentraltest

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