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

# This file describes the browser build for Trichinella spiralis WS245 version

##############################################################################
# download sequence, create UCSC sequence (DONE - 2015-06-24 - Hiram)
    # both the genbank and the WormBase releases are identical
    mkdir -p /hive/data/genomes/triSpi1/genbank
    cd /hive/data/genomes/triSpi1/genbank

    time rsync -L -a -P \
rsync://ftp.ncbi.nlm.nih.gov/genomes/genbank/invertebrate/Trichinella_spiralis/all_assembly_versions/GCA_000181795.2_Trichinella_spiralis-3.7.1/ ./
# sent 394 bytes  received 81911524 bytes  7122775.48 bytes/sec
# total size is 81899323  speedup is 1.00
# real    0m10.375s

    mkdir -p /hive/data/genomes/triSpi1/ws245
    cd /hive/data/genomes/triSpi1/ws245

    wget --no-parent --timestamping -m -nH --cut-dirs=6 \
ftp://ftp.sanger.ac.uk/pub/wormbase/releases/WS245/species/PRJNA12603

    mkdir /hive/data/genomes/triSpi1/ucsc
    cd /hive/data/genomes/triSpi1/ucsc

    zcat ../genbank/GCA_000181795.2_Trichinella_spiralis-3.7.1_assembly_structure/Primary_Assembly/unplaced_scaffolds/FASTA/unplaced.scaf.fna.gz \
        | sed -e 's/.1 Trichinella .*/v1/;' | gzip -c > triSpi1.fa.gz

    zcat ../genbank/GCA_000181795.2_Trichinella_spiralis-3.7.1_assembly_structure/Primary_Assembly/unplaced_scaffolds/AGP/unplaced.scaf.agp.gz \
      | sed -e 's/\.1/v1/;' > triSpi1.agp

    # obtain photo, CDC public domain photo
    mkdir /hive/data/genomes/triSpi1/photo
    cd /hive/data/genomes/triSpi1/photo
    wget --timestamping \
http://www.cdc.gov/dpdx/images/trichinellosis/Trichinella_larvaeF.jpg
    # it is already 300x300 check in to
    #   src/hg/htdocs/images/Trichinella_spiralis.jpg
    # and copy to /usr/local/apache/htdocs/images/Trichinella_spiralis.jpg

#############################################################################
#  Initial database build (DONE - 2015-07-10 - Hiram)

    cd /hive/data/genomes/triSpi1
    cat << '_EOF_' > triSpi1.config.ra
# Config parameters for makeGenomeDb.pl:
db triSpi1
clade worm
genomeCladePriority 68
scientificName Trichinella spiralis
commonName T. spiralis
assemblyDate Jan. 2011
assemblyLabel WashU School of Medicine T. spiralis genome project
assemblyShortLabel WS225/Trichinella_spiralis-3.7.1
orderKey 20030
mitoAcc AF293969.1
fastaFiles /hive/data/genomes/triSpi1/ucsc/triSpi1.fa.gz
agpFiles /hive/data/genomes/triSpi1/ucsc/triSpi1.agp
# qualFiles none
dbDbSpeciesDir worm
photoCreditURL http://www.cdc.gov/dpdx/trichinellosis/gallery.html#larvaebear
photoCreditName Centers for Disease Control and Prevention
ncbiGenomeId 238
ncbiAssemblyId 245458
ncbiAssemblyName WS225/Trichinella_spiralis-3.7.1
ncbiBioProject 12603
genBankAccessionID GCA_000181795.2
taxId 6334
'_EOF_'
    # << happy emacs

    # verify sequence and AGP are OK:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev -fileServer=hgwdev \
         -stop=agp triSpi1.config.ra) > agp.log 2>&1
    # *** All done!  (through the 'agp' step)
    # real    0m20.536s

    # then finish it off:
    time (makeGenomeDb.pl -workhorse=hgwdev -dbHost=hgwdev \
       -fileServer=hgwdev -continue=db triSpi1.config.ra) > db.log 2>&1
    # real    0m51.541s

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

##############################################################################
# cpgIslands on UNMASKED sequence (DONE - 2015-07-10 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/cpgIslandsUnmasked
    cd /hive/data/genomes/triSpi1/bed/cpgIslandsUnmasked

    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku -buildDir=`pwd` \
       -tableName=cpgIslandExtUnmasked \
          -maskedSeq=/hive/data/genomes/triSpi1/triSpi1.unmasked.2bit \
             -workhorse=hgwdev -smallClusterHub=ku triSpi1) > do.log 2>&1
    # real    5m36.177s

    cat fb.triSpi1.cpgIslandExtUnmasked.txt
    # 2557009 bases of 58558192 (4.367%) in intersection

#############################################################################
# cytoBandIdeo - (DONE - 2015-07-10 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/cytoBand
    cd /hive/data/genomes/triSpi1/bed/cytoBand
    makeCytoBandIdeo.csh triSpi1

#########################################################################
# ucscToINSDC table/track (DONE - 2015-07-10 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/ucscToINSDC
    cd /hive/data/genomes/triSpi1/bed/ucscToINSDC

    # find the chrM accession:
    grep chrM ../../*.agp
# chrM    1       16706   2       F       AF293969.1      1       16706   +

    # use that as the second argument here:
    ~/kent/src/hg/utils/automation/ucscToINSDC.sh \
       ../../genbank/GCA_*assembly_structure/Primary_Assembly AF293969.1

    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 *
    #  6864 name.coordinate.tab
    #  6864 ucscToINSDC.bed
    #  6864 ucscToINSDC.txt

    cut -f1 ucscToINSDC.bed | awk '{print length($0)}' | sort -n | tail -1
    # 10
    # use the 10 in this sed
    sed -e "s/21/10/" $HOME/kent/src/hg/lib/ucscToINSDC.sql \
         | hgLoadSqlTab triSpi1 ucscToINSDC stdin ucscToINSDC.bed
    checkTableCoords triSpi1
    # should cover %100 entirely:
    featureBits -countGaps triSpi1 ucscToINSDC
    # 63542128 bases of 63542128 (100.000%) in intersection

#########################################################################
# fixup search rule for assembly track/gold table (DONE - 2015-07-10 - Hiram)
    hgsql -N -e "select frag from gold;" triSpi1 | sort | head -3
ABIR02000001.1
ABIR02000002.1
ABIR02000003.1

    hgsql -N -e "select frag from gold;" triSpi1 | sort | tail -2
ABIR02009267.1
AF293969.1

    # verify this rule will find them all or eliminate them all:
    hgsql -N -e "select frag from gold;" triSpi1 | wc -l
    # 9268

    hgsql -N -e "select frag from gold;" triSpi1 \
      | egrep -e '(ABIR0200|AF293969)[0-9]*(\.1)?' | wc -l
    # 9268

  hgsql -N -e "select frag from gold;" triSpi1 \
      | egrep -v -e '(ABIR0200|AF293969)[0-9]*(\.1)?' | wc -l
    # 0

    # hence, add to trackDb/worm/triSpi1/trackDb.ra
searchTable gold
shortCircuit 1
termRegex (ABIR0200|AF293969)[0-9]*(\.1)?
query select chrom,chromStart,chromEnd,frag from %s where frag like '%s%%'
searchPriority 8

##########################################################################
# running repeat masker (DONE - 2015-07-10 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/repeatMasker
    cd /hive/data/genomes/triSpi1/bed/repeatMasker
    time (doRepeatMasker.pl -buildDir=`pwd` \
        -bigClusterHub=ku -dbHost=hgwdev -workhorse=hgwdev \
        -smallClusterHub=ku triSpi1) > do.log 2>&1
    # real    35m3.117s

    cat faSize.rmsk.txt
    # 63542128 bases (4986988 N's 58555140 real 56390244 upper 2164896 lower)
    #    in 6864 sequences in 1 files
    # Total size: mean 9257.3 sd 248572.9 min 139 (GL624204v1)
    #    max 12041450 (GL622787v1) median 1071
    # %3.41 masked total, %3.70 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 triSpi1 rmsk
    # 2164933 bases of 63542128 (3.407%) in intersection
    # real    0m2.881s

##########################################################################
# running simple repeat (DONE - 2015-07-10 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/simpleRepeat
    cd /hive/data/genomes/triSpi1/bed/simpleRepeat
    time (doSimpleRepeat.pl -buildDir=`pwd` -bigClusterHub=ku \
        -dbHost=hgwdev -workhorse=hgwdev -smallClusterHub=ku \
        triSpi1) > do.log 2>&1
    # real    5m3.124s

    cat fb.simpleRepeat
    # 717851 bases of 58558192 (1.226%) in intersection

    # using the Window Masker result as indicated below

##########################################################################
# CREATE MICROSAT TRACK (DONE - 2015-07-15 - Hiram)
     ssh hgwdev
     mkdir /cluster/data/triSpi1/bed/microsat
     cd /cluster/data/triSpi1/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 triSpi1 microsat microsat.bed
    #	Read 699 elements of size 4 from microsat.bed

##########################################################################
## WINDOWMASKER (DONE - 2015-07-15 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/windowMasker
    cd /hive/data/genomes/triSpi1/bed/windowMasker
    time (doWindowMasker.pl -buildDir=`pwd` -workhorse=hgwdev \
        -dbHost=hgwdev triSpi1) > do.log 2>&1
    # real    4m15.661s

    # Masking statistics
    cat faSize.triSpi1.cleanWMSdust.txt
    # 63542128 bases (4986988 N's 58555140 real 38676889 upper 19878251 lower)
    #    in 6864 sequences in 1 files
    # Total size: mean 9257.3 sd 248572.9 min 139 (GL624204v1)
    #    max 12041450 (GL622787v1) median 1071
    # %31.28 masked total, %33.95 masked real

    cat fb.triSpi1.rmsk.windowmaskerSdust.txt
    # 1699489 bases of 63542128 (2.675%) in intersection

    # using this Window Masker result for final masking::
    cd /hive/data/genomes/triSpi1
    #   you can safely ignore the warning about fields >= 13
    twoBitMask bed/windowMasker/triSpi1.cleanWMSdust.2bit \
       -add bed/simpleRepeat/trfMask.bed  triSpi1.2bit

    # measure the final masking:
    twoBitToFa triSpi1.2bit stdout | faSize stdin > faSize.triSpi1.2bit.txt
    cat faSize.triSpi1.2bit.txt
    # 63542128 bases (4986988 N's 58555140 real 38673309 upper 19881831 lower)
    #   in 6864 sequences in 1 files
    # Total size: mean 9257.3 sd 248572.9 min 139 (GL624204v1)
    #   max 12041450 (GL622787v1) median 1071
    # %31.29 masked total, %33.95 masked real

    # and reset the symlink
    rm /gbdb/triSpi1/triSpi1.2bit
    ln -s /hive/data/genomes/triSpi1/triSpi1.2bit /gbdb/triSpi1/triSpi1.2bit

##########################################################################
# cpgIslands - (DONE - 2015-07-15 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/cpgIslands
    cd /hive/data/genomes/triSpi1/bed/cpgIslands
    time (doCpgIslands.pl -dbHost=hgwdev -bigClusterHub=ku \
      -workhorse=hgwdev -smallClusterHub=ku triSpi1) > do.log 2>&1 &
    # real    15m51.807s

    cat fb.triSpi1.cpgIslandExt.txt
    # 2025345 bases of 58558192 (3.459%) in intersection

#########################################################################
# augustus - (DONE - 2015-07-15 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/augustus
    cd /hive/data/genomes/triSpi1/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 triSpi1) > do.log 2>&1 &
    # real    64m45.505s

    cat fb.triSpi1.augustusGene.txt
    # 8039688 bases of 58558192 (13.729%) in intersection

#########################################################################
# genscan - (DONE - 2015-07-15 - Hiram)
    mkdir /hive/data/genomes/triSpi1/bed/genscan
    cd /hive/data/genomes/triSpi1/bed/genscan
    time (doGenscan.pl -buildDir=`pwd` -workhorse=hgwdev -dbHost=hgwdev \
      -bigClusterHub=ku triSpi1) > do.log 2>&1
    # real    10m7.962s

    cat fb.triSpi1.genscan.txt
    #   4576895 bases of 58558192 (7.816%) in intersection

    cat fb.triSpi1.genscanSubopt.txt
    #   973684 bases of 58558192 (1.663%) in intersection

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

    cd /hive/data/genomes/triSpi1
    # numerator is triSpi1 gapless bases "real" as reported by:
    head -1 faSize.triSpi1.2bit.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/triSpi1
    time blat triSpi1.2bit \
         /dev/null /dev/null -tileSize=11 -makeOoc=jkStuff/triSpi1.11.ooc \
        -repMatch=100
    # Wrote 8527 overused 11-mers to jkStuff/triSpi1.11.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;' triSpi1 \
       | 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 triSpi1 jkStuff/triSpi1.nonBridged.lft \
         -bedFile=jkStuff/triSpi1.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 triSpi1 just before priPac1

# triSpi1 (P. pacificus)
triSpi1.serverGenome = /hive/data/genomes/triSpi1/triSpi1.2bit
triSpi1.clusterGenome = /hive/data/genomes/triSpi1/triSpi1.2bit
triSpi1.ooc = /hive/data/genomes/triSpi1/jkStuff/triSpi1.11.ooc
triSpi1.lift = /hive/data/genomes/triSpi1/jkStuff/triSpi1.nonBridged.lft
triSpi1.refseq.mrna.native.pslCDnaFilter  = ${ordered.refseq.mrna.native.pslCDnaFilter}
triSpi1.refseq.mrna.xeno.pslCDnaFilter    = ${ordered.refseq.mrna.xeno.pslCDnaFilter}
triSpi1.genbank.mrna.native.pslCDnaFilter = ${ordered.genbank.mrna.native.pslCDnaFilter}
triSpi1.genbank.mrna.xeno.pslCDnaFilter   = ${ordered.genbank.mrna.xeno.pslCDnaFilter}
triSpi1.genbank.est.native.pslCDnaFilter  = ${ordered.genbank.est.native.pslCDnaFilter}
triSpi1.refseq.mrna.native.load = yes
triSpi1.refseq.mrna.xeno.load  = yes
triSpi1.refseq.mrna.xeno.loadDesc = yes
# DO NOT NEED genbank.mrna.xeno except for human, mouse
triSpi1.genbank.mrna.xeno.load = no
triSpi1.genbank.est.native.load = yes
triSpi1.genbank.est.native.loadDesc = no
triSpi1.downloadDir = triSpi1
triSpi1.perChromTables = no

    git commit -m "Added triSpi1 - 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 triSpi1
    # logFile: var/build/logs/2015.07.02-11:39:01.triSpi1.initalign.log
    # real    91m20.362s

    #   To re-do, rm the dir first:
    #     /cluster/data/genbank/work/initial.triSpi1

    # load database when finished
    ssh hgwdev
    cd /cluster/data/genbank
    time ./bin/gbDbLoadStep -drop -initialLoad triSpi1
    # logFile: var/dbload/hgwdev/logs/2015.07.06-09:51:31.triSpi1.dbload.log
    # real    18m34.687s

    # enable daily alignment and update of hgwdev
    cd ~/kent/src/hg/makeDb/genbank
    git pull
    # add triSpi1 to:
    #   vi etc/align.dbs etc/hgwdev.dbs
    git commit -m "Added triSpi1 - 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=triSpi1 -tableCoverage all.joiner
    joinerCheck -database=triSpi1 -times all.joiner
    joinerCheck -database=triSpi1 -keys all.joiner

    cd /hive/data/genomes/triSpi1
    time makeDownloads.pl triSpi1 > downloads.log 2>&1
    # real    13m42.027s

    #   now ready for pushQ entry
    mkdir /hive/data/genomes/triSpi1/pushQ
    cd /hive/data/genomes/triSpi1/pushQ
    makePushQSql.pl triSpi1 > triSpi1.pushQ.sql 2> stderr.out
    #   check for errors in stderr.out, some are OK, e.g.:
    # WARNING: hgwdev does not have /gbdb/triSpi1/wib/gc5Base.wib
    # WARNING: hgwdev does not have /gbdb/triSpi1/wib/quality.wib
    # WARNING: hgwdev does not have /gbdb/triSpi1/bbi/qualityBw/quality.bw
    # WARNING: triSpi1 does not have seq
    # WARNING: triSpi1 does not have extFile
    # WARNING: triSpi1 does not have estOrientInfo
    # WARNING: triSpi1 does not have mrnaOrientInfo

    #   copy it to hgwbeta
    scp -p triSpi1.pushQ.sql qateam@hgwbeta:/tmp
    ssh qateam@hgwbeta "./bin/x86_64/hgsql qapushq < /tmp/triSpi1.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="Trichina worm" where name="triSpi1";' hgcentraltest

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