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

# This file describes how the browser for C. malayi WS245 version is built

##############################################################################
# download sequence, create UCSC sequence (DONE - 2015-07-08 - Hiram)
    mkdir -p /hive/data/genomes/strRat1/ws245
    cd /hive/data/genomes/strRat1/ws245

    wget --no-parent --timestamping -m -nH --cut-dirs=6 \
ftp://ftp.sanger.ac.uk/pub/wormbase/releases/WS245/species/PRJEA62033

    mkdir /hive/data/genomes/strRat1/ucsc
    cd /hive/data/genomes/strRat1/ucsc
    # WormBase contig names are of the pattern:
# >RATTI_contig_17 1 1087
# >RATTI_contig_136 1 1168
# >RATTI_contig_139 1 3870
#   ...

    # one duplicate contig removed:
    #   RATTI_contig_75562 and RATTI_contig_75559 are identical

    # remove the size information from the names:
    zcat ../ws245/PRJEA62033/s_ratti.PRJEA62033.WS245.genomic.fa.gz \
        | sed -e 's/ 1.*//;' | gzip -c > strRat1.fa.gz

    hgFakeAgp strRat1.fa.gz strRat1.agp

    # photo from Sanger:
    mkdir /hive/data/genomes/strRat1/photo
    cd /hive/data/genomes/strRat1/photo
    wget --timestamping \
http://www.sanger.ac.uk/resources/downloads/helminths/gfx/strongyloides_ratti.jpg
    # From: Mark Viney/University of Bristol
    #       http://www.bristol.ac.uk/biology/people/mark-e-viney/index.html

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

    cd /hive/data/genomes/strRat1
    cat << '_EOF_' > strRat1.config.ra
# Config parameters for makeGenomeDb.pl:
db strRat1
clade worm
genomeCladePriority 68
scientificName Strongyloides ratti
commonName S. ratti ED321
assemblyDate Apr. 2011
assemblyLabel Wellcome Trust Sanger Institute S. ratti genome project
assemblyShortLabel WS226/S. ratti ED321
orderKey 19050
mitoAcc none
fastaFiles /hive/data/genomes/strRat1/ucsc/strRat1.fa.gz
agpFiles /hive/data/genomes/strRat1/ucsc/strRat1.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 WS226/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 strRat1.config.ra) > agp.log 2>&1
    # *** All done!  (through the 'agp' step)
    # real    0m12.275s

    # then finish it off:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev \
       -fileServer=hgwdev -continue=db strRat1.config.ra) > db.log 2>&1
    # real    0m42.744s

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

##############################################################################
# cpgIslands on UNMASKED sequence (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat1/bed/cpgIslandsUnmasked
    cd /hive/data/genomes/strRat1/bed/cpgIslandsUnmasked

    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku -buildDir=`pwd` \
       -tableName=cpgIslandExtUnmasked \
          -maskedSeq=/hive/data/genomes/strRat1/strRat1.unmasked.2bit \
             -workhorse=hgwdev -smallClusterHub=ku strRat1) > do.log 2>&1
    # real    23m58.455s

    cat fb.strRat1.cpgIslandExtUnmasked.txt
    # 3888494 bases of 52634366 (7.388%) in intersection

#############################################################################
# cytoBandIdeo - (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat1/bed/cytoBand
    cd /hive/data/genomes/strRat1/bed/cytoBand
    makeCytoBandIdeo.csh strRat1

#########################################################################
# ucscToINSDC table/track (TBD - 2015-03-20 - Hiram)
    mkdir /hive/data/genomes/strRat1/bed/ucscToINSDC
    cd /hive/data/genomes/strRat1/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 strRat1 ucscToINSDC stdin ucscToINSDC.bed
    checkTableCoords strRat1
    # should cover %100 entirely:
    featureBits -countGaps strRat1 ucscToINSDC
    # 2053849526 bases of 2053849526 (100.000%) in intersection

#########################################################################
# fixup search rule for assembly track/gold table (TBD - 2015-06-01 - Hiram)
    hgsql -N -e "select frag from gold;" strRat1 | sort | head -3
JF414117.1
Ppa_Contig0_1
Ppa_Contig0_10
[JP][Fp][a4][1_][C4][o1][n1][t7][i\.](g[0-9]*)?(\_[0-9]*)?

    hgsql -N -e "select frag from gold;" strRat1 | sort | tail -2
Ppa_Contig9_98
Ppa_Contig9_99

    # verify this rule will find them all or eliminate them all:
    hgsql -N -e "select frag from gold;" strRat1 | wc -l
    # 28759

    hgsql -N -e "select frag from gold;" strRat1 \
      | egrep -e '[JP][Fp][a4][1_][C4][o1][n1][t7][i\.](g[0-9]*)?(\_[0-9]*)?' \
        | wc -l
    # 28759

  hgsql -N -e "select frag from gold;" strRat1 \
    | egrep -v -e '[JP][Fp][a4][1_][C4][o1][n1][t7][i\.](g[0-9]*)?(\_[0-9]*)?' \
       | wc -l
    # 0

    # hence, add to trackDb/worm/strRat1/trackDb.ra
searchTable gold
shortCircuit 1
termRegex [JP][Fp][a4][1_][C4][o1][n1][t7][i\.](g[0-9]*)?(\_[0-9]*)?
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/strRat1/bed/repeatMasker
    cd /hive/data/genomes/strRat1/bed/repeatMasker
    time (doRepeatMasker.pl -buildDir=`pwd` \
        -bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
        -smallClusterHub=ku strRat1) > do.log 2>&1
    # real    31m2.201s

    cat faSize.rmsk.txt
    # 52634486 bases (11106 N's 52623380 real 50590889 upper 2032491 lower)
    #    in 2183 sequences in 1 files
    # Total size: mean 24111.1 sd 103231.2 min 1001 (RATTI_contig_11208)
    #    max 1635242 (RATTI_contig_75321) median 1771
    # %3.86 masked total, %3.86 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 strRat1 rmsk
    # 2032717 bases of 52634486 (3.862%) in intersection
    # real    0m1.036s

    # 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/strRat1/bed/simpleRepeat
    cd /hive/data/genomes/strRat1/bed/simpleRepeat
    time (doSimpleRepeat.pl -buildDir=`pwd` -bigClusterHub=ku \
        -dbHost=hgwdev -workhorse=hgwdev -smallClusterHub=ku \
        strRat1) > do.log 2>&1
    # real    14m13.669s

    cat fb.simpleRepeat
    # 1545340 bases of 52634366 (2.936%) in intersection

    # using the Window Masker result as indicated below

##########################################################################
# CREATE MICROSAT TRACK (DONE - 2015-07-08 - Hiram)
     ssh hgwdev
     mkdir /cluster/data/strRat1/bed/microsat
     cd /cluster/data/strRat1/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 strRat1 microsat microsat.bed
    #	Read 92 elements of size 4 from microsat.bed

##########################################################################
## WINDOWMASKER (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat1/bed/windowMasker
    cd /hive/data/genomes/strRat1/bed/windowMasker
    time (doWindowMasker.pl -buildDir=`pwd` -workhorse=hgwdev \
        -dbHost=hgwdev strRat1) > do.log 2>&1
    # real    2m16.183s

    # Masking statistics
    cat faSize.strRat1.cleanWMSdust.txt
    # 52634486 bases (11106 N's 52623380 real 24256113 upper 28367267 lower)
    #    in 2183 sequences in 1 files
    # Total size: mean 24111.1 sd 103231.2 min 1001 (RATTI_contig_11208)
    #    max 1635242 (RATTI_contig_75321) median 1771
    # %53.89 masked total, %53.91 masked real

    cat fb.strRat1.rmsk.windowmaskerSdust.txt
    # 1879610 bases of 52634486 (3.571%) in intersection

    # using this Window Masker result for final masking::
    cd /hive/data/genomes/strRat1
    #   you can safely ignore the warning about fields >= 13
    twoBitMask bed/windowMasker/strRat1.cleanWMSdust.2bit \
       -add bed/simpleRepeat/trfMask.bed  strRat1.2bit

    # measure the final masking:
    twoBitToFa strRat1.2bit stdout | faSize stdin > faSize.strRat1.2bit.txt
    cat faSize.strRat1.2bit.txt
    # 52634486 bases (11106 N's 52623380 real 24250940 upper 28372440 lower)
    #    in 2183 sequences in 1 files
    # Total size: mean 24111.1 sd 103231.2 min 1001 (RATTI_contig_11208)
    #    max 1635242 (RATTI_contig_75321) median 1771
    # %53.90 masked total, %53.92 masked real

    # and reset the symlink
    rm /gbdb/strRat1/strRat1.2bit
    ln -s /hive/data/genomes/strRat1/strRat1.2bit /gbdb/strRat1/strRat1.2bit

##########################################################################
# cpgIslands - (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat1/bed/cpgIslands
    cd /hive/data/genomes/strRat1/bed/cpgIslands
    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku \
      -workhorse=hgwdev -smallClusterHub=ku strRat1) > do.log 2>&1 &
    # real    8m10.474s

    cat fb.strRat1.cpgIslandExt.txt
    # 29350 bases of 86344036 (0.034%) in intersection

#########################################################################
# augustus - (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat1/bed/augustus
    cd /hive/data/genomes/strRat1/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 strRat1) > do.log 2>&1
    # real    17m59.242s

    cat fb.strRat1.augustusGene.txt
    # 5462392 bases of 52634366 (10.378%) in intersection

#########################################################################
# genscan - (DONE - 2015-07-08 - Hiram)
    mkdir /hive/data/genomes/strRat1/bed/genscan
    cd /hive/data/genomes/strRat1/bed/genscan
    time (doGenscan.pl -buildDir=`pwd` -workhorse=hgwdev -dbHost=hgwdev \
      -bigClusterHub=ku strRat1) > do.log 2>&1
    # real    3m59.992s

    cat fb.strRat1.genscan.txt
    #   4412238 bases of 52634366 (8.383%) in intersection

    cat fb.strRat1.genscanSubopt.txt
    #   2300611 bases of 52634366 (4.371%) in intersection

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

    cd /hive/data/genomes/strRat1
    # numerator is strRat1 gapless bases "real" as reported by:
    head -1 faSize.strRat1.2bit.txt
# 52634486 bases (11106 N's 52623380 real 24250940 upper 28372440 lower)
#    in 2183 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 \( 52623380 / 2861349177 \) \* 1024
    # ( 52623380 / 2861349177 ) * 1024 = 18.832494

    # ==> use -repMatch=100 since 30 or 50 masks too much
    cd /hive/data/genomes/strRat1
    time blat strRat1.2bit \
         /dev/null /dev/null -tileSize=11 -makeOoc=jkStuff/strRat1.11.ooc \
        -repMatch=100
    # Wrote 7158 overused 11-mers to jkStuff/strRat1.11.ooc
    #  real    0m1.846s

    # there are no non-bridged gaps, do not 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;' strRat1 \
#        | 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 strRat1 jkStuff/strRat1.nonBridged.lft \
         -bedFile=jkStuff/strRat1.nonBridged.bed

    # survey sizes:
    n50.pl chrom.sizes
#       reading: chrom.sizes
#       contig count: 2183, total size: 52634486, one half size: 26317243
# cumulative    N50 count       contig  contig size
# 26053835        38      RATTI_contig_74994      376846
# 26317243 one half size
# 26412864        39      RATTI_contig_74989      359029

#############################################################################
# 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 strRat1 just before priPac1

# strRat1 (P. pacificus)
strRat1.serverGenome = /hive/data/genomes/strRat1/strRat1.2bit
strRat1.clusterGenome = /hive/data/genomes/strRat1/strRat1.2bit
strRat1.ooc = /hive/data/genomes/strRat1/jkStuff/strRat1.11.ooc
strRat1.lift = /hive/data/genomes/strRat1/jkStuff/strRat1.nonBridged.lft
strRat1.refseq.mrna.native.pslCDnaFilter  = ${ordered.refseq.mrna.native.pslCDnaFilter}
strRat1.refseq.mrna.xeno.pslCDnaFilter    = ${ordered.refseq.mrna.xeno.pslCDnaFilter}
strRat1.genbank.mrna.native.pslCDnaFilter = ${ordered.genbank.mrna.native.pslCDnaFilter}
strRat1.genbank.mrna.xeno.pslCDnaFilter   = ${ordered.genbank.mrna.xeno.pslCDnaFilter}
strRat1.genbank.est.native.pslCDnaFilter  = ${ordered.genbank.est.native.pslCDnaFilter}
strRat1.refseq.mrna.native.load = yes
strRat1.refseq.mrna.xeno.load  = yes
strRat1.refseq.mrna.xeno.loadDesc = yes
# DO NOT NEED genbank.mrna.xeno except for human, mouse
strRat1.genbank.mrna.xeno.load = no
strRat1.genbank.est.native.load = yes
strRat1.genbank.est.native.loadDesc = no
strRat1.downloadDir = strRat1
strRat1.perChromTables = no

    git commit -m "Added strRat1 - 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 strRat1
    # logFile: var/build/logs/2015.07.02-11:39:01.strRat1.initalign.log
    # real    91m20.362s

    #   To re-do, rm the dir first:
    #     /cluster/data/genbank/work/initial.strRat1

    # load database when finished
    ssh hgwdev
    cd /cluster/data/genbank
    time ./bin/gbDbLoadStep -drop -initialLoad strRat1
    # logFile: var/dbload/hgwdev/logs/2015.07.06-09:51:31.strRat1.dbload.log
    # real    18m34.687s

    # enable daily alignment and update of hgwdev
    cd ~/kent/src/hg/makeDb/genbank
    git pull
    # add strRat1 to:
    #   vi etc/align.dbs etc/hgwdev.dbs
    git commit -m "Added strRat1 - 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=strRat1 -tableCoverage all.joiner
    joinerCheck -database=strRat1 -times all.joiner
    joinerCheck -database=strRat1 -keys all.joiner

    cd /hive/data/genomes/strRat1
    time makeDownloads.pl strRat1 > downloads.log 2>&1
    # real    13m42.027s

    #   now ready for pushQ entry
    mkdir /hive/data/genomes/strRat1/pushQ
    cd /hive/data/genomes/strRat1/pushQ
    makePushQSql.pl strRat1 > strRat1.pushQ.sql 2> stderr.out
    #   check for errors in stderr.out, some are OK, e.g.:
    # WARNING: hgwdev does not have /gbdb/strRat1/wib/gc5Base.wib
    # WARNING: hgwdev does not have /gbdb/strRat1/wib/quality.wib
    # WARNING: hgwdev does not have /gbdb/strRat1/bbi/qualityBw/quality.bw
    # WARNING: strRat1 does not have seq
    # WARNING: strRat1 does not have extFile
    # WARNING: strRat1 does not have estOrientInfo
    # WARNING: strRat1 does not have mrnaOrientInfo

    #   copy it to hgwbeta
    scp -p strRat1.pushQ.sql qateam@hgwbeta:/tmp
    ssh qateam@hgwbeta "./bin/x86_64/hgsql qapushq < /tmp/strRat1.pushQ.sql"

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

#############################################################################
# LIFTOVER TO strRat2 (DONE - 2015-07-08 - Hiram )
    mkdir /hive/data/genomes/strRat1/bed/blat.strRat2.2015-07-08
    cd /hive/data/genomes/strRat1/bed/blat.strRat2.2015-07-08
    # -debug run to create run dir, preview scripts...
    doSameSpeciesLiftOver.pl \
	-buildDir=`pwd` \
	-bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
  -ooc=/hive/data/genomes/strRat1/jkStuff/strRat1.11.ooc -debug strRat1 strRat2
    # Real run:
    time (doSameSpeciesLiftOver.pl -buildDir=`pwd` \
      -bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
       -ooc=/hive/data/genomes/strRat1/jkStuff/strRat1.11.ooc strRat1 strRat2) \
		> do.log 2>&1
    #	real    3m45.644s

    # verify it works on genome-test

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