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

# This file describes how the browser for Haemonchus contortus
# Haemonchus_contortus_MHco3-2.0
# version is built

##############################################################################
# download sequence, create UCSC sequence (DONE - 2015-06-24 - Hiram)
    mkdir -p /hive/data/genomes/haeCon2/ws245
    cd /hive/data/genomes/haeCon2/ws245

    wget --no-parent --timestamping -m -nH --cut-dirs=6 \
ftp://ftp.sanger.ac.uk/pub/wormbase/releases/WS245/species/PRJEB506

    mkdir /hive/data/genomes/haeCon2/ucsc
    cd /hive/data/genomes/haeCon2/ucsc
    # WormBase contig names are of the pattern:
# >scaffold_1 1 947606
# >scaffold_2 1 768212
# >scaffold_3 1 735772
# >scaffold_4 1 629041
#   ...

    # no conversion of names to maintain equivalence with WormBase:
    zcat ../ws245/PRJEB506/h_contortus.PRJEB506.WS245.genomic.fa.gz \
        | sed -e 's/ 1.*//;' | gzip -c > haeCon2.fa.gz

    hgFakeAgp haeCon2.fa.gz haeCon2.agp

    # correct image from CSIRO:
    mkdir /hive/data/genomes/haeCon2/photo
    cd /hive/data/genomes/haeCon2/photo
    wget --timestamping \
http://www.scienceimage.csiro.au/images/cache/detail/976_0_GA11439.jpg
    convert -quality 80 -geometry 350x350 976_0_GA11439.jpg  Haemonchus_contortus.jpg
    identify Haemonchus_contortus.jpg
    #  Haemonchus_contortus.jpg JPEG 350x279 350x279+0+0

    # reference
    #  http://www.scienceimage.csiro.au/image/10819

#############################################################################
#  Initial database build (DONE - 2015-07-02 - Hiram)

    cd /hive/data/genomes/haeCon2
    cat << '_EOF_' > haeCon2.config.ra
# Config parameters for makeGenomeDb.pl:
db haeCon2
clade worm
# genomeCladePriority 86
scientificName Haemonchus contortus
commonName H. contortus
assemblyDate Jul. 2013
assemblyLabel Sanger Institute
assemblyShortLabel WormBase WS239
orderKey 8004
mitoAcc EU346694.2
fastaFiles /hive/data/genomes/haeCon2/ucsc/haeCon2.fa.gz
agpFiles /hive/data/genomes/haeCon2/ucsc/haeCon2.agp
# qualFiles none
dbDbSpeciesDir worm
photoCreditURL http://www.wormbase.org/species/h_contortus#0--10
photoCreditName Wellcome Trust Sanger Institute H. contortus genome project
ncbiGenomeId 16936
ncbiAssemblyId 64951
ncbiAssemblyName Haemonchus_contortus_MHco3-2.0
ncbiBioProject 194485
genBankAccessionID GCA_000469685.1
taxId 6289
'_EOF_'
    # << happy emacs

    # had to remove duplicate sequences from the original
    # verify sequence and AGP are OK:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev -fileServer=hgwdev \
         -stop=agp haeCon2.config.ra) > agp.log 2>&1
    # *** All done!  (through the 'agp' step)
    # real    0m46.403s

    # then finish it off:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev \
       -fileServer=hgwdev -continue=db haeCon2.config.ra) > db.log 2>&1
    # real    3m34.049s

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

##############################################################################
# cpgIslands on UNMASKED sequence (DONE - 2015-07-02 - Hiram)
    mkdir /hive/data/genomes/haeCon2/bed/cpgIslandsUnmasked
    cd /hive/data/genomes/haeCon2/bed/cpgIslandsUnmasked

    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku -buildDir=`pwd` \
       -tableName=cpgIslandExtUnmasked \
          -maskedSeq=/hive/data/genomes/haeCon2/haeCon2.unmasked.2bit \
             -workhorse=hgwdev -smallClusterHub=ku haeCon2) > do.log 2>&1
    # real    15m12.545s

    cat fb.haeCon2.cpgIslandExtUnmasked.txt
    # 18622402 bases of 346001353 (5.382%) in intersection

#############################################################################
# cytoBandIdeo - (DONE - 2015-07-02 - Hiram)
    mkdir /hive/data/genomes/haeCon2/bed/cytoBand
    cd /hive/data/genomes/haeCon2/bed/cytoBand
    makeCytoBandIdeo.csh haeCon2

#########################################################################
# ucscToINSDC table/track (TBD - 2015-03-20 - Hiram)
    mkdir /hive/data/genomes/haeCon2/bed/ucscToINSDC
    cd /hive/data/genomes/haeCon2/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 haeCon2 ucscToINSDC stdin ucscToINSDC.bed
    checkTableCoords haeCon2
    # should cover %100 entirely:
    featureBits -countGaps haeCon2 ucscToINSDC
    # 2053849526 bases of 2053849526 (100.000%) in intersection

#########################################################################
# fixup search rule for assembly track/gold table (DONE - 2015-07-02 - Hiram)
    hgsql -N -e "select frag from gold;" haeCon2 | sort | head -3
EU346694.2
scaffold_10000_1
scaffold_10001_1

    hgsql -N -e "select frag from gold;" haeCon2 | sort | tail -2
scaffold_9_8
scaffold_9_9

    # verify this rule will find them all or eliminate them all:
    hgsql -N -e "select frag from gold;" haeCon2 | wc -l
    # 54694

    hgsql -N -e "select frag from gold;" haeCon2 \
      | egrep -e '[Es][Uc][adflo0-9]*([_\.])?[0-9]*(\_[0-9]*)?' \
        | wc -l
    # 54694

    hgsql -N -e "select frag from gold;" haeCon2 \
      | egrep -v -e '[Es][Uc][adflo0-9]*([_\.])?[0-9]*(\_[0-9]*)?' \
       | wc -l
    # 0

    # hence, add to trackDb/worm/haeCon2/trackDb.ra
searchTable gold
shortCircuit 1
termRegex [Es][Uc][adflo0-9]*([_\.])?[0-9]*(\_[0-9]*)?
query select chrom,chromStart,chromEnd,frag from %s where frag like '%s%%'
searchPriority 8

##########################################################################
# running repeat masker (DONE - 2015-07-02 - Hiram)
    mkdir /hive/data/genomes/haeCon2/bed/repeatMasker
    cd /hive/data/genomes/haeCon2/bed/repeatMasker
    time (doRepeatMasker.pl -buildDir=`pwd` \
        -bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
        -smallClusterHub=ku haeCon2) > do.log 2>&1
    # real    240m0.986s

    cat faSize.rmsk.txt
    # 369720058 bases (23804161 N's 345915897 real 342326326 upper
    #    3589571 lower) in 23823 sequences in 1 files
    # Total size: mean 15519.5 sd 41044.7 min 101 (scaffold_23862)
    #    max 947606 (scaffold_1) median 1515
    # %0.97 masked total, %1.04 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 haeCon2 rmsk
    # 3590455 bases of 369720058 (0.971%) in intersection
    # real    0m8.681s

    # 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 - 2015-07-02 - Hiram)

    mkdir /hive/data/genomes/haeCon2/bed/simpleRepeat
    cd /hive/data/genomes/haeCon2/bed/simpleRepeat
    time (doSimpleRepeat.pl -buildDir=`pwd` -bigClusterHub=ku \
        -dbHost=hgwdev -workhorse=hgwdev -smallClusterHub=ku \
        haeCon2) > do.log 2>&1
    # real    9m29.002s

    cat fb.simpleRepeat
    # 5523864 bases of 346001353 (1.596%) in intersection

    # using the Window Masker result as indicated below

##########################################################################
# CREATE MICROSAT TRACK (DONE - 2015-07-02 - Hiram)
     ssh hgwdev
     mkdir /cluster/data/haeCon2/bed/microsat
     cd /cluster/data/haeCon2/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 haeCon2 microsat microsat.bed
    #	Read 762 elements of size 4 from microsat.bed

##########################################################################
## WINDOWMASKER (DONE - 2015-07-06 - Hiram)

    mkdir /hive/data/genomes/haeCon2/bed/windowMasker
    cd /hive/data/genomes/haeCon2/bed/windowMasker
    time (doWindowMasker.pl -buildDir=`pwd` -workhorse=hgwdev \
        -dbHost=hgwdev haeCon2) > do.log 2>&1
    # real    26m4.925s

    # Masking statistics
    cat faSize.haeCon2.cleanWMSdust.txt
    # 369720058 bases (23804161 N's 345915897 real 272772728 upper
    #    73143169 lower) in 23823 sequences in 1 files
    # Total size: mean 15519.5 sd 41044.7 min 101 (scaffold_23862)
    #    max 947606 (scaffold_1) median 1515
    # %19.78 masked total, %21.14 masked real

    cat fb.haeCon2.rmsk.windowmaskerSdust.txt
    # 2870130 bases of 369720058 (0.776%) in intersection

    # using this Window Masker result for final masking::
    cd /hive/data/genomes/haeCon2
    #   you can safely ignore the warning about fields >= 13
    twoBitMask bed/windowMasker/haeCon2.cleanWMSdust.2bit \
       -add bed/simpleRepeat/trfMask.bed  haeCon2.2bit

    # measure the final masking:
    twoBitToFa haeCon2.2bit stdout | faSize stdin > faSize.haeCon2.2bit.txt
    cat faSize.haeCon2.2bit.txt
    # 369720058 bases (23804161 N's 345915897 real 272735439 upper
    #    73180458 lower) in 23823 sequences in 1 files
    # Total size: mean 15519.5 sd 41044.7 min 101 (scaffold_23862)
    #    max 947606 (scaffold_1) median 1515
    # %19.79 masked total, %21.16 masked real

    # and reset the symlink
    rm /gbdb/haeCon2/haeCon2.2bit
    ln -s /hive/data/genomes/haeCon2/haeCon2.2bit /gbdb/haeCon2/haeCon2.2bit

##########################################################################
# cpgIslands - (DONE - 2015-07-06 - Hiram)
    mkdir /hive/data/genomes/haeCon2/bed/cpgIslands
    cd /hive/data/genomes/haeCon2/bed/cpgIslands
    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku \
      -workhorse=hgwdev -smallClusterHub=ku haeCon2) > do.log 2>&1 &
    # real    13m56.640s

    cat fb.haeCon2.cpgIslandExt.txt
    # 5472466 bases of 346001353 (1.582%) in intersection

#########################################################################
# augustus - (DONE - 2015-07-06 - Hiram)
    # This is not necessarily correct, the caenorhabditis species setting
    # is not ideal here:
    mkdir /hive/data/genomes/haeCon2/bed/augustus
    cd /hive/data/genomes/haeCon2/bed/augustus
    time (doAugustus.pl -buildDir=`pwd` -bigClusterHub=ku \
      -species=caenorhabditis -dbHost=hgwdev \
         -workhorse=hgwdev haeCon2) > do.log 2>&1
    # real    70m9.005s

    cat fb.haeCon2.augustusGene.txt
    # 36765684 bases of 346001353 (10.626%) in intersection

#########################################################################
# genscan - (DONE - 2015-07-06 - Hiram)
    mkdir /hive/data/genomes/haeCon2/bed/genscan
    cd /hive/data/genomes/haeCon2/bed/genscan
    time (doGenscan.pl -buildDir=`pwd` -workhorse=hgwdev -dbHost=hgwdev \
      -bigClusterHub=ku haeCon2) > do.log 2>&1
    # real    24m20.289s

    cat fb.haeCon2.genscan.txt
    #   20302013 bases of 346001353 (5.868%) in intersection

    cat fb.haeCon2.genscanSubopt.txt
    #   21532138 bases of 346001353 (6.223%) in intersection

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

    cd /hive/data/genomes/haeCon2
    # numerator is haeCon2 gapless bases "real" as reported by:
    head -1 faSize.haeCon2.2bit.txt
# 369720058 bases (23804161 N's 345915897 real 272735439 upper 73180458 lower)
# in 23823 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 \( 345915897 / 2861349177 \) \* 1024
    # ( 345915897 / 2861349177 ) * 1024 = 123.794006

    # ==> use -repMatch=100 according to size scaled down from 1024 for human.
    #   and rounded down to nearest 50
    cd /hive/data/genomes/haeCon2
    time blat haeCon2.2bit \
         /dev/null /dev/null -tileSize=11 -makeOoc=jkStuff/haeCon2.11.ooc \
        -repMatch=100
    # Wrote 15059 overused 11-mers to jkStuff/haeCon2.11.ooc
    # real    0m6.125s

    # 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;' haeCon2 \
       | ave -tableOut -col=7 stdin
# min Q1 median Q3 max mean N sum stddev
# 50702 51980 52642 74987 83718 61113.4 17 1.03893e+06 11299.2

    # note the minimum non-bridged gap size is 50,702

    gapToLift -verbose=2 -minGap=50000 haeCon2 jkStuff/haeCon2.nonBridged.lft \
         -bedFile=jkStuff/haeCon2.nonBridged.bed

    # survey sizes:
    n50.pl chrom.sizes
#       reading: chrom.sizes
#       contig count: 23823, total size: 369720058, one half size: 184860029
# cumulative    N50 count       contig  contig size
# 184836276       1149    scaffold_1149   83305
# 184860029 one half size
# 184919575       1150    scaffold_1150   83299

#############################################################################
# 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 haeCon2 just before priPac1

# haeCon2 (P. pacificus)
haeCon2.serverGenome = /hive/data/genomes/haeCon2/haeCon2.2bit
haeCon2.clusterGenome = /hive/data/genomes/haeCon2/haeCon2.2bit
haeCon2.ooc = /hive/data/genomes/haeCon2/jkStuff/haeCon2.11.ooc
haeCon2.lift = /hive/data/genomes/haeCon2/jkStuff/haeCon2.nonBridged.lft
haeCon2.refseq.mrna.native.pslCDnaFilter  = ${ordered.refseq.mrna.native.pslCDnaFilter}
haeCon2.refseq.mrna.xeno.pslCDnaFilter    = ${ordered.refseq.mrna.xeno.pslCDnaFilter}
haeCon2.genbank.mrna.native.pslCDnaFilter = ${ordered.genbank.mrna.native.pslCDnaFilter}
haeCon2.genbank.mrna.xeno.pslCDnaFilter   = ${ordered.genbank.mrna.xeno.pslCDnaFilter}
haeCon2.genbank.est.native.pslCDnaFilter  = ${ordered.genbank.est.native.pslCDnaFilter}
haeCon2.refseq.mrna.native.load = yes
haeCon2.refseq.mrna.xeno.load  = yes
haeCon2.refseq.mrna.xeno.loadDesc = yes
# DO NOT NEED genbank.mrna.xeno except for human, mouse
haeCon2.genbank.mrna.xeno.load = no
haeCon2.genbank.est.native.load = yes
haeCon2.genbank.est.native.loadDesc = no
haeCon2.downloadDir = haeCon2
haeCon2.perChromTables = no

    git commit -m "Added haeCon2 - 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 haeCon2
    # logFile: var/build/logs/2015.07.02-11:39:01.haeCon2.initalign.log
    # real    1651m57.350s

    #   To re-do, rm the dir first:
    #     /cluster/data/genbank/work/initial.haeCon2

    # load database when finished
    ssh hgwdev
    cd /cluster/data/genbank
    time ./bin/gbDbLoadStep -drop -initialLoad haeCon2
    # logFile: var/dbload/hgwdev/logs/2015.06.09-14:34:12.haeCon2.dbload.log
    # real    128m32.009s

    # enable daily alignment and update of hgwdev
    cd ~/kent/src/hg/makeDb/genbank
    git pull
    # add haeCon2 to:
    #   vi etc/align.dbs etc/hgwdev.dbs
    git commit -m "Added haeCon2 - Tibetan frog refs #15056" \
      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=haeCon2 -tableCoverage all.joiner
    joinerCheck -database=haeCon2 -times all.joiner
    joinerCheck -database=haeCon2 -keys all.joiner

    cd /hive/data/genomes/haeCon2
    time makeDownloads.pl haeCon2 > downloads.log 2>&1
    # real    13m42.027s

    #   now ready for pushQ entry
    mkdir /hive/data/genomes/haeCon2/pushQ
    cd /hive/data/genomes/haeCon2/pushQ
    makePushQSql.pl haeCon2 > haeCon2.pushQ.sql 2> stderr.out
    #   check for errors in stderr.out, some are OK, e.g.:
    # WARNING: hgwdev does not have /gbdb/haeCon2/wib/gc5Base.wib
    # WARNING: hgwdev does not have /gbdb/haeCon2/wib/quality.wib
    # WARNING: hgwdev does not have /gbdb/haeCon2/bbi/qualityBw/quality.bw
    # WARNING: haeCon2 does not have seq
    # WARNING: haeCon2 does not have extFile
    # WARNING: haeCon2 does not have estOrientInfo
    # WARNING: haeCon2 does not have mrnaOrientInfo

    #   copy it to hgwbeta
    scp -p haeCon2.pushQ.sql qateam@hgwbeta:/tmp
    ssh qateam@hgwbeta "./bin/x86_64/hgsql qapushq < /tmp/haeCon2.pushQ.sql"

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

#########################################################################
# LIFTOVER TO haeCon1 (DONE - 2015-07-07 - Hiram )
    mkdir /hive/data/genomes/haeCon2/bed/blat.haeCon1.2015-07-07
    cd /hive/data/genomes/haeCon2/bed/blat.haeCon1.2015-07-07
    # -debug run to create run dir, preview scripts...
    doSameSpeciesLiftOver.pl \
	-buildDir=`pwd` \
	-bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
  -ooc=/hive/data/genomes/haeCon2/jkStuff/haeCon2.11.ooc -debug haeCon2 haeCon1
    # Real run:
    time (doSameSpeciesLiftOver.pl -buildDir=`pwd` \
      -bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
       -ooc=/hive/data/genomes/haeCon2/jkStuff/haeCon2.11.ooc haeCon2 haeCon1) \
		> do.log 2>&1
    #	real    11m17.699s
    # verify it works on genome-test

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

    hgsql -e 'update dbDb set organism="Barber pole worm" where name="haeCon2";' hgcentraltest

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