#############################################################################
## 20-Way Multiz (DONE - 2015-04-13 - Hiram)
    ssh hgwdev
    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way
    cd /hive/data/genomes/tarSyr2/bed/multiz20way

    # from the 183-way in the source tree, select out the 20 used here:
    /cluster/bin/phast/tree_doctor \
        --prune-all-but hg38,panTro4,panPan1,gorGor3,ponAbe2,nomLeu3,rheMac3,macFas5,papAnu2,chlSab2,nasLar1,rhiRox1,calJac3,saiBol1,tarSyr2,micMur1,otoGar3,tupBel1,mm10,canFam3 \
        /cluster/home/hiram/kent/src/hg/utils/phyloTrees/183way.nh \
          > tarSyr2.20way.nh

    # using TreeGraph2 tree editor on the Mac, rearrange to get tarSyr2
    # at the top:

    #	what that looks like:
 ~/kent/src/hg/utils/phyloTrees/asciiTree.pl tarSyr2.20way.nh | sed -e 's/^/# /;'
# (((((tarSyr2:0.1114,
#     ((calJac3:0.03,
#      saiBol1:0.01035):0.01965,
#     (((nasLar1:0.0015,
#       rhiRox1:0.0015):0.018,
#      (chlSab2:0.027,
#      (papAnu2:0.008042,
#      (rheMac3:0.004991,
#      macFas5:0.004991):0.0030):0.01061):0.0050):0.02,
#     (((((hg38:0.00655,
#         panTro4:0.00684):0.00122,
#        panPan1:0.00784):0.003,
#       gorGor3:0.008964):0.009693,
#      ponAbe2:0.01894):0.003471,
#     nomLeu3:0.02227):0.01204):0.02183):0.05209):0.02052,
#    (micMur1:0.0856,
#    otoGar3:0.1194):0.02052):0.013494,
#   tupBel1:0.19114):0.002,
#  mm10:0.356483):0.020593,
# canFam3:0.165928);

    # extract species list from that .nh file
    sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \
        tarSyr2.20way.nh | xargs echo | sed 's/ //g; s/,/ /g' \
        | sed 's/[()]//g; s/,/ /g' | tr '[ ]' '[\n]' > species.list.txt

    # construct db to name translation list:
    cat species.list.txt | while read DB
do
hgsql -N -e "select name,organism from dbDb where name=\"${DB}\";" hgcentraltest
done | sed -e "s/\t/->/; s/ /_/g;" | sed -e 's/$/;/' | sed -e 's/\./_/g' \
        | sed -e 's/-nosed/_nosed/; s/-eating/_eating/;' > db.to.name.txt

    # construct a common name .nh file:
    /cluster/bin/phast/tree_doctor --rename \
    "`cat db.to.name.txt`" tarSyr2.20way.nh | sed -e 's/00*)/)/g; s/00*,/,/g' \
       | $HOME/kent/src/hg/utils/phyloTrees/asciiTree.pl /dev/stdin \
         > tarSyr2.20way.commonNames.nh
    cat tarSyr2.20way.commonNames.nh | sed -e 's/^/# /;'
# (((((Tarsier:0.1114,
#     ((Marmoset:0.03,
#      Squirrel_monkey:0.01035):0.01965,
#     (((Proboscis_monkey:0.0015,
#       Golden_snub_nosed_monkey:0.0015):0.018,
#      (Green_monkey:0.027,
#      (Baboon:0.008042,
#      (Rhesus:0.004991,
#      Crab_eating_macaque:0.004991):0.003):0.01061):0.005):0.02,
#     (((((Human:0.00655,
#         Chimp:0.00684):0.00122,
#        Bonobo:0.00784):0.003,
#       Gorilla:0.008964):0.009693,
#      Orangutan:0.01894):0.003471,
#     Gibbon:0.02227):0.01204):0.02183):0.05209):0.02052,
#    (Mouse_lemur:0.0856,
#    Bushbaby:0.1194):0.02052):0.013494,
#   Tree_shrew:0.19114):0.002,
#  Mouse:0.356483):0.020593,
# Dog:0.165928);

#	Use this specification in the phyloGif tool:
#	http://genome.ucsc.edu/cgi-bin/phyloGif
#	to obtain a png image for src/hg/htdocs/images/phylo/tarSyr2_20way.png

    ~/kent/src/hg/utils/phyloTrees/asciiTree.pl tarSyr2.20way.nh > t.nh
    ~/kent/src/hg/utils/phyloTrees/scientificNames.sh t.nh \
       | $HOME/kent/src/hg/utils/phyloTrees/asciiTree.pl /dev/stdin \
          > tarSyr2.20way.scientificNames.nh
    rm -f t.nh
    cat tarSyr2.20way.scientificNames.nh | sed -e 's/^/# /;'
# (((((Tarsius_syrichta:0.1114,
#     ((Callithrix_jacchus:0.03,
#      Saimiri_boliviensis:0.01035):0.01965,
#     (((Nasalis_larvatus:0.0015,
#       Rhinopithecus_roxellana:0.0015):0.018,
#      (Chlorocebus_sabaeus:0.027,
#      (Papio_anubis:0.008042,
#      (Macaca_mulatta:0.004991,
#      Macaca_fascicularis:0.004991):0.003):0.01061):0.005):0.02,
#     (((((Homo_sapiens:0.00655,
#         Pan_troglodytes:0.00684):0.00122,
#        Pan_paniscus:0.00784):0.003,
#       Gorilla_gorilla_gorilla:0.008964):0.009693,
#      Pongo_pygmaeus_abelii:0.01894):0.003471,
#     Nomascus_leucogenys:0.02227):0.01204):0.02183):0.05209):0.02052,
#    (Microcebus_murinus:0.0856,
#    Otolemur_garnettii:0.1194):0.02052):0.013494,
#   Tupaia_belangeri:0.19114):0.002,
#  Mus_musculus:0.356483):0.020593,
# Canis_lupus_familiaris:0.165928);

    /cluster/bin/phast/all_dists tarSyr2.20way.nh | grep tarSyr2 \
        | sed -e "s/tarSyr2.//" | sort -k2n > 20way.distances.txt
    #	Use this output to create the table below
    cat 20way.distances.txt | sed -e 's/^/# /;'
# saiBol1       0.193490
# calJac3       0.213140
# gorGor3       0.219488
# nomLeu3       0.219630
# ponAbe2       0.219771
# hg38  0.221294
# panPan1       0.221364
# panTro4       0.221584
# nasLar1       0.224820
# rhiRox1       0.224820
# macFas5       0.228921
# rheMac3       0.228921
# papAnu2       0.228972
# chlSab2       0.237320
# micMur1       0.238040
# otoGar3       0.271840
# canFam3       0.333935
# tupBel1       0.336554
# mm10  0.503897

    cat << '_EOF_' > sizeStats.pl
#!/usr/bin/env perl

use strict;
use warnings;

open (FH, "<20way.distances.txt") or
        die "can not read 20way.distances.txt";

my $count = 0;
while (my $line = <FH>) {
    chomp $line;
    my ($D, $dist) = split('\s+', $line);
    my $chain = "chain" . ucfirst($D);
    my $B="/hive/data/genomes/tarSyr2/bed/lastz.$D/fb.tarSyr2." .
        $chain . "Link.txt";
    my $chainLinkMeasure =
        `awk '{print \$5}' ${B} 2> /dev/null | sed -e "s/(//; s/)//"`;
    chomp $chainLinkMeasure;
    $chainLinkMeasure = 0.0 if (length($chainLinkMeasure) < 1);
    $chainLinkMeasure =~ s/\%//;
    my $swapFile="/hive/data/genomes/${D}/bed/lastz.tarSyr2/fb.${D}.chainTarSyr2Link.txt";
    my $swapMeasure = "N/A";
    if ( -s $swapFile ) {
	$swapMeasure =
	    `awk '{print \$5}' ${swapFile} 2> /dev/null | sed -e "s/(//; s/)//"`;
	chomp $swapMeasure;
	$swapMeasure = 0.0 if (length($swapMeasure) < 1);
	$swapMeasure =~ s/\%//;
    }
    my $orgName=
    `hgsql -N -e "select organism from dbDb where name='$D';" hgcentraltest`;
    chomp $orgName;
    if (length($orgName) < 1) {
        $orgName="N/A";
    }
    ++$count;
    printf "# %02d  %.4f (%% %06.3f) (%% %06.3f) - %s %s\n", $count, $dist,
        $chainLinkMeasure, $swapMeasure, $orgName, $D;
}
close (FH);
'_EOF_'
    # << happy emacs
    chmod +x ./sizeStats.pl
    ./sizeStats.pl
#

#	If you can fill in all the numbers in this table, you are ready for
#	the multiple alignment procedure

#       featureBits chainLink measures
#               chainLink
#  N distance  on tarSyr2  on other     other species
# 01  0.1935 (% 48.208) (% 59.766) - Squirrel monkey saiBol1
# 02  0.2131 (% 48.629) (% 55.800) - Marmoset calJac3
# 03  0.2195 (% 50.262) (% 56.297) - Gorilla gorGor3
# 04  0.2196 (% 50.545) (% 56.775) - Gibbon nomLeu3
# 05  0.2198 (% 51.302) (% 55.431) - Orangutan ponAbe2
# 06  0.2213 (% 52.305) (% 56.022) - Human hg38
# 07  0.2214 (% 52.289) (% 59.834) - Bonobo panPan1
# 08  0.2216 (% 51.284) (% 56.630) - Chimp panTro4
# 09  0.2248 (% 47.164) (% 61.165) - Proboscis monkey nasLar1
# 10  0.2248 (% 51.539) (% 56.545) - Golden snub-nosed monkey rhiRox1
# 11  0.2289 (% 51.814) (% 57.176) - Crab-eating macaque macFas5
# 12  0.2289 (% 50.710) (% 59.428) - Rhesus rheMac3
# 13  0.2290 (% 51.652) (% 56.346) - Baboon papAnu2
# 14  0.2373 (% 51.691) (% 58.191) - Green monkey chlSab2
# 15  0.2380 (% 38.087) (% 64.649) - Mouse lemur micMur1
# 16  0.2718 (% 44.814) (% 59.044) - Bushbaby otoGar3
# 17  0.3339 (% 44.555) (% 57.932) - Dog canFam3
# 18  0.3366 (% 31.268) (% 47.221) - Tree shrew tupBel1
# 19  0.5039 (% 26.433) (% 32.301) - Mouse mm10

# None of this concern for distances matters in building the first step, the
# maf files.  The distances will be better calibrated later.

    # create species list and stripped down tree for autoMZ
    sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \
	tarSyr2.20way.nh | xargs echo | sed 's/ //g; s/,/ /g' > tree.nh

    sed 's/[()]//g; s/,/ /g' tree.nh > species.list
    # tarSyr2 calJac3 saiBol1 nasLar1 rhiRox1 chlSab2 papAnu2 rheMac3 macFas5
    # hg38 panTro4 panPan1 gorGor3 ponAbe2 nomLeu3 micMur1 otoGar3 tupBel1
    # mm10 canFam3

    #	bash shell syntax here ...
    cd /hive/data/genomes/tarSyr2/bed/multiz20way
    export H=/hive/data/genomes/tarSyr2/bed
    mkdir mafLinks
    # good assemblies can use syntenic net:
    #  hg38 panTro4 nomLeu3 ponAbe2 chlSab2 macFas5 rheMac3 papAnu2 calJac3
    #  mm10 canFam3
    for G in hg38 panTro4 nomLeu3 ponAbe2 chlSab2 macFas5 rheMac3 papAnu2 calJac3 mm10 canFam3
    do
      mkdir mafLinks/$G
      echo ln -s ${H}/lastz.$G/axtChain/tarSyr2.${G}.synNet.maf.gz ./mafLinks/$G
      ln -s ${H}/lastz.$G/axtChain/tarSyr2.${G}.synNet.maf.gz ./mafLinks/$G
    done

    # other assemblies using recip best net:
    #  saiBol1 panPan1 gorGor3 nasLar1 rhiRox1 micMur1 otoGar3 tupBel1
    for G in  saiBol1 panPan1 gorGor3 nasLar1 rhiRox1 micMur1 otoGar3 tupBel1
    do
      mkdir mafLinks/$G
      echo ln -s ${H}/lastz.$G/mafRBestNet/tarSyr2.${G}.rbest.maf.gz ./mafLinks/$G
      ln -s ${H}/lastz.$G/mafRBestNet/tarSyr2.${G}.rbest.maf.gz ./mafLinks/$G
    done

    # verify the symLinks are good:
    ls -ogrtL mafLinks/*/* | sed -e 's/^/# /; s/-rw-rw-r-- 1//;'
#  1053246893 Dec 13 22:14 mafLinks/calJac3/tarSyr2.calJac3.synNet.maf.gz
#  1122406104 Dec 13 22:35 mafLinks/hg38/tarSyr2.hg38.synNet.maf.gz
#  1082953585 Dec 13 22:44 mafLinks/ponAbe2/tarSyr2.ponAbe2.synNet.maf.gz
#  1079125054 Dec 13 22:45 mafLinks/nomLeu3/tarSyr2.nomLeu3.synNet.maf.gz
#  1076470498 Dec 14 23:19 mafLinks/gorGor3/tarSyr2.gorGor3.rbest.maf.gz
#   826258737 Dec 15 00:05 mafLinks/micMur1/tarSyr2.micMur1.rbest.maf.gz
#  1099163205 Dec 15 20:36 mafLinks/panTro4/tarSyr2.panTro4.synNet.maf.gz
#  1123592074 Dec 15 22:03 mafLinks/papAnu2/tarSyr2.papAnu2.synNet.maf.gz
#  1113325538 Dec 16 04:10 mafLinks/chlSab2/tarSyr2.chlSab2.synNet.maf.gz
#  1092254730 Dec 16 04:14 mafLinks/rheMac3/tarSyr2.rheMac3.synNet.maf.gz
#  1110954577 Dec 16 04:29 mafLinks/macFas5/tarSyr2.macFas5.synNet.maf.gz
#   982870684 Dec 16 10:01 mafLinks/otoGar3/tarSyr2.otoGar3.rbest.maf.gz
#  1048154412 Dec 16 12:03 mafLinks/saiBol1/tarSyr2.saiBol1.rbest.maf.gz
#  1136796591 Dec 16 17:43 mafLinks/panPan1/tarSyr2.panPan1.rbest.maf.gz
#  1015087549 Dec 16 19:16 mafLinks/nasLar1/tarSyr2.nasLar1.rbest.maf.gz
#  1110293706 Feb 24 16:17 mafLinks/rhiRox1/tarSyr2.rhiRox1.rbest.maf.gz
#   718392380 Mar 31 14:09 mafLinks/tupBel1/tarSyr2.tupBel1.rbest.maf.gz
#   601154099 Apr  2 13:19 mafLinks/mm10/tarSyr2.mm10.synNet.maf.gz
#  1019074133 Apr  6 00:47 mafLinks/canFam3/tarSyr2.canFam3.synNet.maf.gz


    # split the maf files into a set of hashed named files
    # this hash named split keeps the same chr/contig names in the same
    # named hash file.
    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way/mafSplit
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/mafSplit
    time for D in `sed -e "s/tarSyr2 //" ../species.list`
do
    echo "${D}"
    mkdir $D
    cd $D
    echo "mafSplit -byTarget -useHashedName=8 /dev/null . ../../mafLinks/${D}/*.maf.gz"
    mafSplit -byTarget -useHashedName=8 /dev/null . \
	../../mafLinks/${D}/*.maf.gz
    cd ..
done
    # real    23m28.569s

    # construct a list of all possible maf file names.
    # they do not all exist in each of the species directories
    find . -type f | wc -l
    # 4864
    find . -type f | grep ".maf$" | xargs -L 1 basename | sort -u > maf.list
    wc -l maf.list
    # 256 maf.list

    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way/splitRun
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/splitRun
    mkdir maf run
    cd run
    mkdir penn
    cp -p /cluster/bin/penn/multiz.2009-01-21_patched/multiz penn
    cp -p /cluster/bin/penn/multiz.2009-01-21_patched/maf_project penn
    cp -p /cluster/bin/penn/multiz.2009-01-21_patched/autoMZ penn

    #	verify the db and pairs settings are correct
    cat > autoMultiz.csh << '_EOF_'
#!/bin/csh -ef
set db = tarSyr2
set c = $1
set result = $2
set run = `/bin/pwd`
set tmp = /dev/shm/$db/multiz.$c
set pairs = /hive/data/genomes/tarSyr2/bed/multiz20way/mafSplit
/bin/rm -fr $tmp
/bin/mkdir -p $tmp
/bin/cp -p ../../tree.nh ../../species.list $tmp
pushd $tmp > /dev/null
foreach s (`/bin/sed -e "s/$db //" species.list`)
    set in = $pairs/$s/$c
    set out = $db.$s.sing.maf
    if (-e $in.gz) then
        /bin/zcat $in.gz > $out
        if (! -s $out) then
            echo "##maf version=1 scoring=autoMZ" > $out
        endif
    else if (-e $in) then
        /bin/ln -s $in $out
    else
        echo "##maf version=1 scoring=autoMZ" > $out
    endif
end
set path = ($run/penn $path); rehash
$run/penn/autoMZ + T=$tmp E=$db "`cat tree.nh`" $db.*.sing.maf $c \
        > /dev/null
popd > /dev/null
/bin/rm -f $result
/bin/cp -p $tmp/$c $result
/bin/rm -fr $tmp
'_EOF_'
# << happy emacs
    chmod +x autoMultiz.csh

    cat  << '_EOF_' > template
#LOOP
./autoMultiz.csh $(file1) {check out line+ /hive/data/genomes/tarSyr2/bed/multiz20way/splitRun/maf/$(root1).maf}
#ENDLOOP
'_EOF_'
# << happy emacs

    ln -s ../../mafSplit/maf.list maf.list
    ssh ku
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/splitRun/run
    gensub2 maf.list single template jobList
    para create jobList
    para try ... check ... push ... etc...
# Completed: 256 of 256 jobs
# CPU time in finished jobs:    1490665s   24844.42m   414.07h   17.25d  0.047 y
# IO & Wait Time:                  1865s      31.08m     0.52h    0.02d  0.000 y
# Average job time:                5830s      97.17m     1.62h    0.07d
# Longest finished job:           10512s     175.20m     2.92h    0.12d
# Submission to last job:         10559s     175.98m     2.93h    0.12d


    # combine into one file  (the 1>&2 redirect sends the echo to stderr)
    cd /hive/data/genomes/tarSyr2/bed/multiz20way
    head -1 splitRun/maf/020.maf > multiz20way.maf
    time for F in splitRun/maf/*.maf
do
    echo "${F}" 1>&2
    egrep -v "^#" ${F}
done >> multiz20way.maf
    # real    4m11.298s

    tail -1 splitRun/maf/020.maf >> multiz20way.maf
# -rw-rw-r--  1 55291768197 Apr 13 21:30 multiz20way.maf

    # Load into database
    ssh hgwdev
    cd /hive/data/genomes/tarSyr2/bed/multiz20way
    mkdir /gbdb/tarSyr2/multiz20way
    ln -s `pwd`/multiz20way.maf /gbdb/tarSyr2/multiz20way
    cd /dev/shm
    time hgLoadMaf tarSyr2 multiz20way
# Loaded 28622703 mafs in 1 files from /gbdb/tarSyr2/multiz20way
#  real    13m40.020s

    time hgLoadMafSummary -verbose=2 -minSize=30000 \
	-mergeGap=1500 -maxSize=200000 tarSyr2 multiz20waySummary \
	/gbdb/tarSyr2/multiz20way/multiz20way.maf
# Created 676756 summary blocks from 57801948 components and 28622703 mafs from /gbdb/tarSyr2/multiz20way/multiz20way.maf
#  real    14m1.079s


# -rw-rw-r--  1 1465456549 Apr 13 21:41 multiz20way.tab
# -rw-rw-r--  1   32338422 Apr 13 22:33 multiz20waySummary.tab

    wc -l multiz20way*.tab
#  28622703 multiz20way.tab
#    676756 multiz20waySummary.tab

    rm multiz20way*.tab

##############################################################################
# GAP ANNOTATE MULTIZ7WAY MAF AND LOAD TABLES (DONE - 2015-04-14 - Hiram)
    # mafAddIRows has to be run on single chromosome maf files, it does not
    #	function correctly when more than one reference sequence
    #	are in a single file.  Need to split of the maf file into individual
    #   maf files
    mkdir -p /hive/data/genomes/tarSyr2/bed/multiz20way/anno/mafSplit
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/anno/mafSplit

    time mafSplit -outDirDepth=2 -byTarget -useFullSequenceName \
        /dev/null . ../../multiz20way.maf
    #   real    18m58.525s

    find . -type f | wc -l
    #   89197

    # check for N.bed files everywhere:
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/anno
    for DB in `cat ../species.list`
do
    if [ ! -s /hive/data/genomes/${DB}/${DB}.N.bed ]; then
        echo "MISS: ${DB}"
#         cd /hive/data/genomes/${DB}
#         twoBitInfo -nBed ${DB}.2bit ${DB}.N.bed
    else
        echo "  OK: ${DB}"
    fi
done

    cd /hive/data/genomes/tarSyr2/bed/multiz20way/anno
    for DB in `cat ../species.list`
do
    echo "${DB} "
    ln -s  /hive/data/genomes/${DB}/${DB}.N.bed ${DB}.bed
    echo ${DB}.bed  >> nBeds
    ln -s  /hive/data/genomes/${DB}/chrom.sizes ${DB}.len
    echo ${DB}.len  >> sizes
done
    # make sure they all are successful symLinks:
    ls -ogrtL

    screen -S gapAnno      # use a screen to control this longish job
    ssh ku
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/anno
    mkdir result
    find ./mafSplit -type d | sed -e 's#./mafSplit/##' | while read D
do
    echo mkdir -p result/${D}
    mkdir -p result/${D}
done
    cat << '_EOF_' > template
#LOOP
mafAddIRows -nBeds=nBeds mafSplit/$(path1) /hive/data/genomes/tarSyr2/tarSyr2.2bit {check out exists+ result/$(path1)}
#ENDLOOP
'_EOF_'
    # << happy emacs

    find ./mafSplit -type f | sed -e 's#^./mafSplit/##' > maf.list
    gensub2 maf.list single template jobList
    # limit jobs on a node with the ram=32g requirement because they go fast
    para -ram=32g create jobList
    para try ... check ... push ...
# Completed: 89197 of 89197 jobs
# CPU time in finished jobs:     169379s    2822.98m    47.05h    1.96d  0.005 y
# IO & Wait Time:                229156s    3819.27m    63.65h    2.65d  0.007 y
# Average job time:                   4s       0.07m     0.00h    0.00d
# Longest finished job:              13s       0.22m     0.00h    0.00d
# Submission to last job:          1746s      29.10m     0.48h    0.02d

    # verify all result files have some content, look for 0 size files:
    find ./result -type f -size 0
    # should see none
    # or in this manner:
    find ./result -type f | xargs ls -og | sort -k3nr | tail

    # combine into one file  (the 1>&2 redirect sends the echo to stderr)
    head -q -n 1  result/8/4/ABRT02305379v1.maf > tarSyr2.20way.maf
    time find ./result -type f | while read F
do
    echo "${F}" 1>&2
    grep -h -v "^#" ${F}
done >> tarSyr2.20way.maf
    # real    29m25.870s

    #	these maf files do not have the end marker, this does nothing:
    #	tail -q -n 1 result/8/4/ABRT02305379v1.maf >> tarSyr2.20way.maf
    # How about an official end marker:
    echo "##eof maf" >> tarSyr2.20way.maf
    ls -og
# -rw-rw-r--  1 84345995664 Apr 14 09:27 tarSyr2.20way.maf

    du -hsc tarSyr2.20way.maf
    # 79G     tarSyr2.20way.maf

    # construct symlinks to get the individual maf files into gbdb:
    rm /gbdb/tarSyr2/multiz20way/multiz20way.maf   # remove previous results
    ln -s `pwd`/tarSyr2.20way.maf /gbdb/tarSyr2/multiz20way/multiz20way.maf

    # Load into database
    cd /dev/shm
    time hgLoadMaf -pathPrefix=/gbdb/tarSyr2/multiz20way tarSyr2 multiz20way
# Loaded 30369211 mafs in 1 files from /gbdb/tarSyr2/multiz20way
# real    19m43.000s

    time hgLoadMafSummary -verbose=2 -minSize=30000 \
	-mergeGap=1500 -maxSize=200000 tarSyr2 multiz20waySummary \
        /gbdb/tarSyr2/multiz20way/multiz20way.maf
    # Created 676756 summary blocks from 57801948 components and 30369211 mafs from /gbdb/tarSyr2/multiz20way/multiz20way.maf
    # real    18m49.166s

    # -rw-rw-r--  1 1556576627 Apr 14 09:45 multiz20way.tab
    # -rw-rw-r--  1   33691934 Apr 14 10:18 multiz20waySummary.tab

    rm multiz20way*.tab

######################################################################
# MULTIZ7WAY MAF FRAMES (DONE - 2015-04-14 - Hiram)
    ssh hgwdev
    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way/frames
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/frames
#   survey all the genomes to find out what kinds of gene tracks they have
    cat << '_EOF_' > showGenes.csh
#!/bin/csh -fe
foreach db (`cat ../species.list`)
    echo -n "${db}: "
    set tables = `hgsql $db -N -e "show tables like '%Gene%'"`
    foreach table ($tables)
        if ($table == "ensGene" || $table == "refGene" || \
           $table == "mgcGenes" || $table == "knownGene" || \
           $table == "xenoRefGene" ) then
           set count = `hgsql $db -N -e "select count(*) from $table"`
            echo -n "${table}: ${count}, "
        endif
    end
    set orgName = `hgsql hgcentraltest -N -e \
            "select scientificName from dbDb where name='$db'"`
    set orgId = `hgsql hg19 -N -e \
            "select id from organism where name='$orgName'"`
    if ($orgId == "") then
        echo "Mrnas: 0"
    else
        set count = `hgsql hg19 -N -e "select count(*) from gbCdnaInfo where organism=$orgId"`
        echo "Mrnas: ${count}"
    endif
end
'_EOF_'
    # << happy emacs
    chmod +x ./showGenes.csh
    time ./showGenes.csh
# tarSyr2: xenoRefGene: 317648, Mrnas: 8
# calJac3: ensGene: 55116, refGene: 220, xenoRefGene: 314108, Mrnas: 294478
# saiBol1: xenoRefGene: 444473, Mrnas: 85
# nasLar1: xenoRefGene: 325305, Mrnas: 4
# rhiRox1: xenoRefGene: 331524, Mrnas: 11
# chlSab2: xenoRefGene: 221155, Mrnas: 37897
# papAnu2: ensGene: 29030, refGene: 493, xenoRefGene: 306504, Mrnas: 146334
# rheMac3: refGene: 6533, xenoRefGene: 285257, Mrnas: 443919
# macFas5: refGene: 2270, xenoRefGene: 285932, Mrnas: 177465
# hg38: ensGene: 208239, knownGene: 104178, mgcGenes: 34081, refGene: 57740, xenoRefGene: 178096, Mrnas: 10985957
# panTro4: ensGene: 29160, refGene: 2685, xenoRefGene: 290723, Mrnas: 11222
# panPan1: xenoRefGene: 397450, Mrnas: 566
# gorGor3: ensGene: 35410, xenoRefGene: 351847, Mrnas: 1
# ponAbe2: ensGene: 29447, refGene: 3567, xenoRefGene: 299670, Mrnas: 0
# nomLeu3: xenoRefGene: 198058, Mrnas: 49
# micMur1: ensGene: 37458, xenoRefGene: 555344, Mrnas: 59
# otoGar3: ensGene: 28565, xenoRefGene: 420000, Mrnas: 13
# tupBel1: ensGene: 34727, xenoRefGene: 663425, Mrnas: 2506
# mm10: ensGene: 103734, knownGene: 63244, mgcGenes: 26768, refGene: 34922, xenoRefGene: 166911, Mrnas: 5226373
# canFam3: ensGene: 29884, refGene: 2160, xenoRefGene: 256615, Mrnas: 387766

# real    5m46.618s

    # from that summary, use these gene sets:
    # knownGene - hg38 mm10
    # ensGene - canFam3 panTro4 gorGor3 ponAbe2 papAnu2 calJac3 micMur1
    #     otoGar3 tupBel1
    # refGene - macFas5 rheMac3
    # xenoRefGene - tarSyr2
    # no annotation: nasLar1 rhiRox1 panPan1 nomLeu3 chlSab2 saiBol1

    #   1. knownGene: hg38
    for DB in hg38 mm10
do
    hgsql -N -e "select name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds from knownGene" ${DB} \
      | genePredSingleCover stdin stdout | gzip -2c \
        > genes/${DB}.gp.gz
    genePredCheck -db=${DB} genes/${DB}.gp.gz
done
    # checked: 21887 failed: 0
    # checked: 21036 failed: 0

    #   2. ensGene: canFam3 panTro4 gorGor3 ponAbe2 papAnu2 calJac3 micMur1
    #               otoGar3 tupBel1
    for DB in canFam3 panTro4 gorGor3 ponAbe2 papAnu2 calJac3 micMur1 \
       otoGar3 tupBel1
do
hgsql -N -e "select
name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds
from ensGene" ${DB} \
      | genePredSingleCover stdin stdout | gzip -2c \
        > /dev/shm/${DB}.tmp.gz
    mv /dev/shm/${DB}.tmp.gz genes/$DB.gp.gz
    echo -n "# ${DB}: "
    genePredCheck -db=${DB} genes/${DB}.gp.gz
done
# canFam3: checked: 19507 failed: 0
# panTro4: checked: 18657 failed: 0
# gorGor3: checked: 20758 failed: 0
# ponAbe2: checked: 20220 failed: 0
# papAnu2: checked: 18904 failed: 0
# calJac3: checked: 20827 failed: 0
# micMur1: checked: 28641 failed: 0
# otoGar3: checked: 19472 failed: 0
# tupBel1: checked: 29256 failed: 0

    #   3. refGene for macFas5 rheMac3
    for DB in macFas5 rheMac3
do
hgsql -N -e "select * from refGene" ${DB} | cut -f2- \
      | genePredSingleCover stdin stdout | gzip -2c \
        > /dev/shm/${DB}.tmp.gz
    mv /dev/shm/${DB}.tmp.gz genes/$DB.gp.gz
    echo -n "# ${DB}: "
    genePredCheck -db=${DB} genes/${DB}.gp.gz
done
# macFas5: checked: 2196 failed: 0
# rheMac3: checked: 5801 failed: 0

    #   4. xenoRefGene for tarSyr2
    for DB in tarSyr2
do
    hgsql -N -e "select
name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds
from xenoRefGene" ${DB} \
      | genePredSingleCover stdin stdout | gzip -2c \
        > /dev/shm/${DB}.tmp.gz
    mv /dev/shm/${DB}.tmp.gz genes/$DB.gp.gz
    echo -n "# ${DB}: "
    genePredCheck -db=${DB} genes/${DB}.gp.gz
done
# tarSyr2: checked: 29314 failed: 0

    # verify counts for genes are reasonable:
    for T in genes/*.gz
do
    echo -n "# $T: "
    zcat $T | cut -f1 | sort | uniq -c | wc -l
done
# genes/calJac3.gp.gz: 20827
# genes/canFam3.gp.gz: 19507
# genes/gorGor3.gp.gz: 20758
# genes/hg38.gp.gz: 21887
# genes/macFas5.gp.gz: 2120
# genes/micMur1.gp.gz: 16240
# genes/mm10.gp.gz: 22159
# genes/otoGar3.gp.gz: 19472
# genes/panTro4.gp.gz: 18657
# genes/papAnu2.gp.gz: 18904
# genes/ponAbe2.gp.gz: 20220
# genes/rheMac3.gp.gz: 5626
# genes/tarSyr2.gp.gz: 24868
# genes/tupBel1.gp.gz: 15407

    time (cat ../anno/tarSyr2.20way.maf \
	| genePredToMafFrames tarSyr2 stdin stdout \
          `egrep -v "nasLar1|rhiRox1|panPan1|nomLeu3|chlSab2|saiBol1" \
            ../species.list.txt | xargs echo \
              | sed -e "s#\([a-zA-Z0-9]*\)#\1 genes/\1.gp.gz#g;"` \
		| gzip > multiz20wayFrames.bed.gz)
    #   real    18m8.453s

    # verify there are frames on everything, should be 14 species:
    zcat multiz20wayFrames.bed.gz | awk '{print $4}' | sort | uniq -c
# 223819 calJac3
# 233806 canFam3
# 221546 gorGor3
# 220701 hg38
#  16309 macFas5
# 200742 micMur1
# 211533 mm10
# 214022 otoGar3
# 205286 panTro4
# 209513 papAnu2
# 219206 ponAbe2
#  45987 rheMac3
# 215621 tarSyr2
# 195293 tupBel1

    #   load the resulting file
    ssh hgwdev
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/frames
    time hgLoadMafFrames tarSyr2 multiz20wayFrames multiz20wayFrames.bed.gz
    #   real    0m26.783s

    time featureBits -countGaps tarSyr2 multiz20wayFrames
    # 42787229 bases of 3453864774 (1.239%) in intersection
    # real    1m54.920s

    #   enable the trackDb entries:
# frames multiz20wayFrames
# irows on
    #   appears to work OK

#########################################################################
# Phylogenetic tree from 20-way (DONE - 2015-04-14 - Hiram)
    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way/4d
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/4d

    # using the xenoRefGenes
    hgsql -N -e "select
name,chrom,strand,txStart,txEnd,cdsStart,cdsEnd,exonCount,exonStarts,exonEnds
from xenoRefGene" tarSyr2 \
      | genePredSingleCover stdin stdout > /dev/shm/tarSyr2.tmp.gp
    mv /dev/shm/tarSyr2.tmp.gp tarSyr2.xenoRefGene.gp
    genePredCheck -db=tarSyr2 tarSyr2.xenoRefGene.gp
    # checked: 29314 failed: 0

    genePredSingleCover tarSyr2.xenoRefGene.gp stdout \
       | sort > tarSyr2.xenoRefGeneNR.gp
    genePredCheck -db=tarSyr2 tarSyr2.xenoRefGeneNR.gp
    # checked: 29314 failed: 0

    # the annotated maf is:
    og ../anno/tarSyr2.20way.maf
# -rw-rw-r-- 1 84345995664 Apr 14 09:27 ../anno/tarSyr2.20way.maf

    mkdir annoSplit
    cd annoSplit
    time mafSplit -verbose=2 -outDirDepth=2 -byTarget -useFullSequenceName \
	/dev/null . ../../anno/tarSyr2.20way.maf
    # real    32m20.418s

    find . -type f | wc -l
    #   89197
    ssh ku
    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way/4d/run
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/4d/run
    mkdir ../mfa

    # newer versions of msa_view have a slightly different operation
    # the sed of the gp file inserts the reference species in the chr name
    cat << '_EOF_' > 4d.csh
#!/bin/csh -fe
set PHASTBIN = /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin
set GP = tarSyr2.xenoRefGeneNR.gp
set r = "/hive/data/genomes/tarSyr2/bed/multiz20way"
set c = $1:r
set infile = $r/4d/annoSplit/$2
set outDir = $r/4d/mfa/$3:h
set outfile = $r/4d/mfa/$3
/bin/mkdir -p $outDir
cd /dev/shm
/bin/awk -v C=$c '$2 == C {print}' $r/4d/$GP | sed -e "s/\t$c\t/\ttarSyr2.$c\t/" > $c.gp
set NL=`wc -l $c.gp| gawk '{print $1}'`
echo $NL
if ("$NL" != "0") then
    $PHASTBIN/msa_view --4d --features $c.gp -i MAF $infile -o SS > $c.ss
    $PHASTBIN/msa_view -i SS --tuple-size 1 $c.ss > $outfile
else
    echo "" > $outfile
endif
/bin/rm -f /dev/shm/$c.gp /dev/shm/$c.ss
'_EOF_'
    # << happy emacs
    chmod +x 4d.csh

    find ../annoSplit -type f | sed -e "s#../annoSplit/##" > maf.list

    cat << '_EOF_' > template
#LOOP
4d.csh $(file1) $(path1) {check out line+ ../mfa/$(dir1)/$(dir2)$(root1).mfa}
#ENDLOOP
'_EOF_'
    # << happy emacs

    gensub2 maf.list single template jobList
    para create jobList
    para try ... check
    para time
# Completed: 88756 of 89197 jobs
# Crashed: 441 jobs
# CPU time in finished jobs:       5002s      83.37m     1.39h    0.06d  0.000 y
# IO & Wait Time:                273236s    4553.93m    75.90h    3.16d  0.009 y
# Average job time:                   3s       0.05m     0.00h    0.00d
# Longest finished job:               8s       0.13m     0.00h    0.00d
# Submission to last job:          1291s      21.52m     0.36h    0.01d

    # Not all results have contents, that is OK

    # combine mfa files
    ssh hgwdev
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/4d
    # remove the broken empty files, size 0 and size 1:
    find ./mfa -type f -size 0 | xargs rm -f
    # sometimes this doesn't work, don't know why
    find ./mfa -type f -size 1 | xargs rm -f
    # when it doesn't, use this empty list procedure
    find ./mfa -type f | xargs ls -og | awk '$3 < 2' | awk '{print $NF}' \
        > empty.list
    cat empty.list | xargs rm -f
    # see what is left:
    ls -ogrt mfa/*/*/*.mfa | sort -k3nr | wc
    #       9256   64792  524134

    # want comma-less species.list
    time /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/msa_view \
	--aggregate "`cat ../species.list`" mfa/*/*/*.mfa | sed s/"> "/">"/ \
	    > 4d.all.mfa
    # real    1m34.617s

    # check they are all in there:
    grep "^>" 4d.all.mfa | wc -l
    # 20
    grep "^>" 4d.all.mfa | sed -e 's/^/# /;'
# >tarSyr2
# >calJac3
# >saiBol1
# >nasLar1
# >rhiRox1
# >chlSab2
# >papAnu2
# >rheMac3
# >macFas5
# >hg38
# >panTro4
# >panPan1
# >gorGor3
# >ponAbe2
# >nomLeu3
# >micMur1
# >otoGar3
# >tupBel1
# >mm10
# >canFam3

    sed 's/[a-z][a-z]*_//g; s/:[0-9\.][0-9\.]*//g; s/;//; /^ *$/d' \
	../tarSyr2.20way.nh | xargs echo | sed -e 's/ //g' > tree_commas.nh
    # tree_commas.nh looks like:
    # (((((tarSyr2,((calJac3,saiBol1),(((nasLar1,rhiRox1),(chlSab2,(papAnu2,(rheMac3,macFas5)))),(((((hg38,panTro4),panPan1),gorGor3),ponAbe2),nomLeu3)))),(micMur1,otoGar3)),tupBel1),mm10),canFam3)

    # use phyloFit to create tree model (output is phyloFit.mod)
    time /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/phyloFit \
	    --EM --precision MED --msa-format FASTA --subst-mod REV \
		--tree tree_commas.nh 4d.all.mfa
    #  real    3m36.651s


    mv phyloFit.mod all.mod

    grep TREE all.mod
# TREE:
# (((((tarSyr2:0.102759,((calJac3:0.0255607,saiBol1:0.0242179):0.026921,
# (((nasLar1:0.00671106,rhiRox1:0.00510423):0.00871252,(chlSab2:0.00943875,
# (papAnu2:0.00584608,(rheMac3:0.00235964,
# macFas5:0.00211514):0.00374983):0.00282721):0.00425991):0.0154827,
# (((((hg38:0.0073972,panTro4:0.00239992):0.00199447,
# panPan1:0.00217438):0.00419837,gorGor3:0.0072327):0.00690608,
# ponAbe2:0.0140231):0.00246106,
# nomLeu3:0.0164131):0.00834414):0.0152742):0.045029):0.00928622,
# (micMur1:0.0677891,otoGar3:0.0921232):0.0253299):0.0133967,
# tupBel1:0.134026):0.0139795,mm10:0.227865):0.0603994,canFam3:0.0603994);

    # compare these calculated lengths to the tree extracted from 183way:
    grep TREE all.mod | sed -e 's/TREE: //' \
       | /cluster/bin/phast/all_dists /dev/stdin | grep tarSyr2 \
          | sed -e "s/tarSyr2.//;"  | sort > new.dists
    grep TREE /hive/data/genomes/hg38/bed/multiz20way.0/4d/all.mod \
       | sed -e 's/TREE: //' \
       | /cluster/bin/phast/all_dists /dev/stdin | grep tarSyr2 \
          | sed -e "s/tarSyr2.//;"  | sort > hg38.dists
    /cluster/bin/phast/all_dists ../tarSyr2.20way.nh | grep tarSyr2 \
        | sed -e "s/tarSyr2.//;" | sort > old.dists
     # printing out the 'new', the 'old' the 'difference' and percent difference
    join new.dists old.dists | awk '{
  printf "#\t%s\t%8.6f\t%8.6f\t%8.6f\t%8.6f\n", $1, $2, $3, $2-$3, 100*($2-$3)/$3 }' \
      | sort -k3n
#       panPan1 0.187146        0.221364        -0.034218       -15.457798
#       nomLeu3 0.187819        0.219630        -0.031811       -14.483905
#       ponAbe2 0.187891        0.219771        -0.031880       -14.506009
#       gorGor3 0.188006        0.219488        -0.031482       -14.343381
#       panTro4 0.189366        0.221584        -0.032218       -14.539858
#       papAnu2 0.191478        0.228972        -0.037494       -16.374928
#       macFas5 0.191497        0.228921        -0.037424       -16.347998
#       rheMac3 0.191741        0.228921        -0.037180       -16.241411
#       chlSab2 0.192244        0.237320        -0.045076       -18.993764
#       rhiRox1 0.192362        0.224820        -0.032458       -14.437328
#       nasLar1 0.193968        0.224820        -0.030852       -13.722978
#       hg38    0.194364        0.221294        -0.026930       -12.169331
#       saiBol1 0.198927        0.193490        0.005437        2.809964
#       calJac3 0.200270        0.213140        -0.012870       -6.038285
#       micMur1 0.205164        0.238040        -0.032876       -13.811124
#       otoGar3 0.229498        0.271840        -0.042342       -15.576074
#       tupBel1 0.259468        0.336554        -0.077086       -22.904497
#       canFam3 0.260220        0.333935        -0.073715       -22.074655
#       mm10    0.367286        0.503897        -0.136611       -27.110898

     # compare this tree with the one generated from the hg38 20-way
    grep TREE all.mod | sed -e 's/TREE: //' \
     | /cluster/bin/phast/all_dists /dev/stdin | grep hg38 \
        | sed -e "s/hg38.//" | sort > tarSyr2.hg38.dists
    grep TREE /hive/data/genomes/hg38/bed/multiz20way/4d/all.mod \
       | sed -e 's/TREE: //' \
       | /cluster/bin/phast/all_dists /dev/stdin | grep hg38 \
          | sed -e "s/hg38.//;"  | sort > hg38.hg38.dists
    # appears to be systematically shorter branch lengths with this tarSyr2
    # reference, all by a similar percentage:
    # table headings:
#       species  tarSyr2         hg38         tarSyr2-hg38       percent
#                dist est.     dist est.         difference      difference
    join tarSyr2.hg38.dists hg38.hg38.dists | awk '{
  printf "#\t%s\t%8.6f\t%8.6f\t%8.6f\t%8.6f\n", $1, $2, $3, $2-$3, 100*($2-$3)/$3 }' \
      | sort -k3n
#       panTro4 0.009797        0.013354        -0.003557       -26.636214
#       panPan1 0.011566        0.014412        -0.002846       -19.747433
#       gorGor3 0.020823        0.026014        -0.005191       -19.954640
#       ponAbe2 0.034519        0.045236        -0.010717       -23.691308
#       nomLeu3 0.039370        0.052852        -0.013482       -25.508968
#       papAnu2 0.059717        0.081458        -0.021741       -26.689828
#       macFas5 0.059736        0.080661        -0.020925       -25.941905
#       rheMac3 0.059981        0.081710        -0.021729       -26.592828
#       chlSab2 0.060483        0.081818        -0.021335       -26.076169
#       rhiRox1 0.060601        0.081305        -0.020704       -25.464609
#       nasLar1 0.062208        0.081742        -0.019534       -23.897140
#       saiBol1 0.097714        0.133183        -0.035469       -26.631777
#       calJac3 0.099057        0.135678        -0.036621       -26.991111
#       micMur1 0.194010        0.263345        -0.069335       -26.328580
#       tarSyr2 0.194364        0.266293        -0.071929       -27.011224
#       otoGar3 0.218344        0.299131        -0.080787       -27.007231
#       tupBel1 0.248313        0.337009        -0.088696       -26.318585
#       canFam3 0.249066        0.343407        -0.094341       -27.472067
#       mm10    0.356132        0.498858        -0.142726       -28.610546

#########################################################################
# phastCons 20-way (DONE - 2015-04-15 - Hiram)
    # split 20way mafs into 10M chunks and generate sufficient statistics
    # files for # phastCons
    ssh ku
    mkdir -p /hive/data/genomes/tarSyr2/bed/multiz20way/cons/SS
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/cons/SS
    mkdir result done

    cat << '_EOF_' > mkSS.csh
#!/bin/csh -ef
set d = $1
set c = $2
set doneDir = done/$d
set MAF = /hive/data/genomes/tarSyr2/bed/multiz20way/anno/result/$d/$c.maf
set WINDOWS = /hive/data/genomes/tarSyr2/bed/multiz20way/cons/SS/result/$d/$c
set WC = `cat $MAF | wc -l`
set NL = `grep "^#" $MAF | wc -l`
if ( -s $3 ) then
    exit 0
endif
if ( -s $3.running ) then
    exit 0
endif

/bin/mkdir -p $doneDir
/bin/date >> $3.running

/bin/rm -fr $WINDOWS
/bin/mkdir -p $WINDOWS
pushd $WINDOWS > /dev/null
if ( $WC != $NL ) then
/cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/msa_split \
    $MAF -i MAF -o SS -r $WINDOWS/$c -w 10000000,0 -I 1000 -B 5000
endif
popd > /dev/null
/bin/date >> $3
/bin/rm -f $3.running
'_EOF_'
    # << happy emacs
    chmod +x mkSS.csh

    cat << '_EOF_' > template
#LOOP
mkSS.csh $(dir1) $(root1) {check out line+ done/$(dir1)/$(root1)}
#ENDLOOP
'_EOF_'
    # << happy emacs

    find ../../anno/result -type f | sed -e "s#../../anno/result/##" > maf.list

    ssh ku
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/cons/SS

    gensub2 maf.list single template jobList
    # beware overwhelming the cluster with these quick high I/O jobs
    para create jobList
    para try ... check ... etc
    para -maxJob=64 push
# Completed: 89176 of 89197 jobs
# Crashed: 21 jobs
# CPU time in finished jobs:       7761s     129.36m     2.16h    0.09d  0.000 y
# IO & Wait Time:                233500s    3891.66m    64.86h    2.70d  0.007 y
# Average job time:                   3s       0.05m     0.00h    0.00d
# Longest finished job:              10s       0.17m     0.00h    0.00d
# Submission to last job:          6211s     103.52m     1.73h    0.07d

    # the 21 jobs that failed are due to tiny one base MAF files as input
    # and there are lots of results that are completely empty

    find ./result -type f | wc -l
    #	 19343

    # Run phastCons
    #	This job is I/O intensive in its output files, beware where this
    #	takes place or do not run too many at once.
    ssh ku
    mkdir -p /hive/data/genomes/tarSyr2/bed/multiz20way/cons/run.cons
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/cons/run.cons

    #	This is setup for multiple runs based on subsets, but only running
    #   the 'all' subset here.
    #   It triggers off of the current working directory
    #	$cwd:t which is the "grp" in this script.  Running:
    #	all and vertebrates

    cat << '_EOF_' > doPhast.csh
#!/bin/csh -fe
set PHASTBIN = /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin
set c = $1
set d = $2
set f = $3
set len = $4
set cov = $5
set rho = $6
set grp = $cwd:t
set cons = /hive/data/genomes/tarSyr2/bed/multiz20way/cons
set tmp = $cons/tmp/${d}_${c}
mkdir -p $tmp
set ssSrc = $cons/SS/result
set useGrp = "$grp.mod"
if (-s $cons/$grp/$grp.non-inf) then
  ln -s $cons/$grp/$grp.mod $tmp
  ln -s $cons/$grp/$grp.non-inf $tmp
  ln -s $ssSrc/$d/$f $tmp
else
  ln -s $ssSrc/$d/$f $tmp
  ln -s $cons/$grp/$grp.mod $tmp
endif
pushd $tmp > /dev/null
if (-s $grp.non-inf) then
  $PHASTBIN/phastCons $f $useGrp \
    --rho $rho --expected-length $len --target-coverage $cov --quiet \
    --not-informative `cat $grp.non-inf` \
    --seqname $c --idpref $c --most-conserved $c.bed --score > $c.pp
else
  $PHASTBIN/phastCons $f $useGrp \
    --rho $rho --expected-length $len --target-coverage $cov --quiet \
    --seqname $c --idpref $c --most-conserved $c.bed --score > $c.pp
endif
popd > /dev/null
mkdir -p pp/$d bed/$d
sleep 4
touch pp/$d bed/$d
rm -f pp/$d/$c.pp
rm -f bed/$d/$c.bed
mv $tmp/$c.pp pp/$d
mv $tmp/$c.bed bed/$d
rm -fr $tmp
rmdir --ignore-fail-on-non-empty $cons/tmp/$d:h
'_EOF_'
    # << happy emacs
    chmod +x doPhast.csh

    #	this template will serve for all runs
    #	root1 == chrom name, file1 == ss file name without .ss suffix
    cat << '_EOF_' > template
#LOOP
../run.cons/doPhast.csh $(root1) $(dir1) $(file1) 45 0.3 0.3 {check out line+ pp/$(dir1)/$(root1).pp}
#ENDLOOP
'_EOF_'
    # << happy emacs

    find ../SS/result -type f | sed -e "s#../SS/result/##" > ss.list
    wc -l ss.list
    #	19343 ss.list

    # Create parasol batch and run it
    # run for all species
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/cons
    mkdir -p all
    cd all
    #	Using the .mod tree
    cp -p ../../4d/all.mod ./all.mod

    gensub2 ../run.cons/ss.list single ../run.cons/template jobList
    para -ram=32g create jobList
    para try ... check ...
    para push
# Completed: 19343 of 19343 jobs
# CPU time in finished jobs:      13357s     222.61m     3.71h    0.15d  0.000 y
# IO & Wait Time:                127906s    2131.77m    35.53h    1.48d  0.004 y
# Average job time:                   7s       0.12m     0.00h    0.00d
# Longest finished job:              20s       0.33m     0.01h    0.00d
# Submission to last job:          3519s      58.65m     0.98h    0.04d

    # create Most Conserved track
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/cons/all
    time cut -f1 ../../../../chrom.sizes | while read C
do
    ls -d bed/?/?/${C} 2> /dev/null | while read D
    do
        echo ${D}/${C}*.bed 1>&2
        cat ${D}/${C}*.bed
    done | sort -k1,1 -k2,2n \
    | awk '{printf "%s\t%d\t%d\tlod=%d\t%s\n", "'${C}'", $2, $3, $5, $5;}'
done > tmpMostConserved.bed
    # real    46m0.788s

    time /cluster/bin/scripts/lodToBedScore tmpMostConserved.bed \
         > mostConserved.bed
    # real    0m6.335s

    # -rw-rw-r--  1 24980443 Apr 14 18:49 mostConserved.bed

    # load into database
    ssh hgwdev
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/cons/all
    time hgLoadBed tarSyr2 phastConsElements20way mostConserved.bed
    # Read 701406 elements of size 5 from mostConserved.bed
    # real    0m7.088s

    # on human we often try for 5% overall cov, and 70% CDS cov
    # most bets are off here for that goal, these alignments are too few
    #	and too far between
    #	--rho 0.3 --expected-length 45 --target-coverage 0.3
    time featureBits tarSyr2 -enrichment xenoRefGene:cds phastConsElements20way
    # xenoRefGene:cds 1.086%, phastConsElements20way 1.890%, both 0.523%,
    # cover 48.15%, enrich 25.47x

    # Create merged posterier probability file and wiggle track data files
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/cons/all
    mkdir downloads

    # the third sed fixes the chrom names, removing the partition extensions
    time (find ./pp -type f | sed -e "s#^./##; s#\.# d #g; s#-# m #;" \
	| sort -k1,1 -k3,3n | sed -e "s# d #.#g; s# m #-#g;" | xargs cat \
	| sed -e 's/\.[0-9][0-9]*-[0-9][0-9]* start/ start/' \
        | gzip -c > downloads/phastCons20way.wigFix.gz)
    #   real    23m21.925s
# -rw-rw-r-- 1 1727340003 Apr 14 19:13 phastCons20way.wigFix.gz

    # check integrity of data with wigToBigWig
    time (zcat downloads/phastCons20way.wigFix.gz \
	| wigToBigWig -verbose=2 stdin /hive/data/genomes/tarSyr2/chrom.sizes \
	    phastCons20way.bw) > bigWig.log 2>&1
    tail bigWig.log
    # pid=29517: VmPeak:    22583404 kB
    # real    35m34.029s

    bigWigInfo phastCons20way.bw | sed -e 's/^/# /;'
# version: 4
# isCompressed: yes
# isSwapped: 0
# primaryDataSize: 2,588,525,608
# primaryIndexSize: 99,299,640
# zoomLevels: 10
# chromCount: 19343
# basesCovered: 2,059,732,828
# mean: 0.050953
# min: 0.000000
# max: 1.000000
# std: 0.171558

    #	encode those files into wiggle data
    time (zcat downloads/phastCons20way.wigFix.gz \
	| wigEncode stdin phastCons20way.wig phastCons20way.wib)
    # Converted stdin, upper limit 1.00, lower limit 0.00
    #  real    10m13.824s

    du -hsc *.wi?
    #  2.0G    phastCons20way.wib
    #  309M    phastCons20way.wig

    # Load gbdb and database with wiggle.
    ln -s `pwd`/phastCons20way.wib /gbdb/tarSyr2/multiz20way/phastCons20way.wib
    time hgLoadWiggle -pathPrefix=/gbdb/tarSyr2/multiz20way \
	tarSyr2 phastCons20way phastCons20way.wig
    #   real    0m36.479s

    # use to set trackDb.ra entries for wiggle min and max
    # and verify table is loaded correctly

    wigTableStats.sh tarSyr2 phastCons20way
# db.table          min max mean       count sumData
# tarSyr2.phastCons20way  0 1 0.0509527 2059732828 1.04949e+08
#     stdDev  viewLimits
#   0.171558 viewLimits=0:0.908742


    #  Create histogram to get an overview of all the data
    time hgWiggle -doHistogram -db=tarSyr2 \
	-hBinSize=0.001 -hBinCount=1000 -hMinVal=0.0 -verbose=2 \
	    phastCons20way > histogram.data 2>&1
    #	real    7m28.605s

    #	create plot of histogram:

    cat << '_EOF_' | gnuplot > histo.png
set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff font \
"/usr/share/fonts/default/Type1/n022004l.pfb"
set size 1.4, 0.8
set key left box
set grid noxtics
set grid ytics
set title " Tarsier tarSyr2 Histogram phastCons20way track"
set xlabel " phastCons20way score"
set ylabel " Relative Frequency"
set y2label " Cumulative Relative Frequency (CRF)"
set y2range [0:1]
set y2tics
set yrange [0:0.02]

plot "histogram.data" using 2:5 title " RelFreq" with impulses, \
        "histogram.data" using 2:7 axes x1y2 title " CRF" with lines
'_EOF_'
    #	<< happy emacs

    display histo.png &

#########################################################################
# phyloP for 20-way (DONE - 2015-04-15 - Hiram)
    # run phyloP with score=LRT
    ssh ku
    mkdir /cluster/data/tarSyr2/bed/multiz20way/consPhyloP
    cd /cluster/data/tarSyr2/bed/multiz20way/consPhyloP

    mkdir run.phyloP
    cd run.phyloP
    # Adjust model file base composition background and rate matrix to be
    # representative of the chromosomes in play
    grep BACKGROUND ../../4d/all.mod | awk '{printf "%0.3f\n", $3 + $4}'
    #	0.508
    /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin/modFreqs \
	../../4d/all.mod 0.508 > all.mod
    # verify, the BACKGROUND should now be paired up:
    grep BACK all.mod
    #   BACKGROUND: 0.240500 0.259500 0.259500 0.240500 

    cat << '_EOF_' > doPhyloP.csh
#!/bin/csh -fe
set PHASTBIN = /cluster/bin/phast.build/cornellCVS/phast.2010-12-30/bin
set f = $1
set d = $f:h
set file1 = $f:t
set out = $2
set cName = $f:t:r
set grp = $cwd:t
set cons = /hive/data/genomes/tarSyr2/bed/multiz20way/consPhyloP
set tmp = $cons/tmp/$grp/$f
/bin/rm -fr $tmp
/bin/mkdir -p $tmp
set ssSrc = "/hive/data/genomes/tarSyr2/bed/multiz20way/cons/SS/result/$f"
set useGrp = "$grp.mod"
/bin/ln -s $cons/run.phyloP/$grp.mod $tmp
pushd $tmp > /dev/null
$PHASTBIN/phyloP --method LRT --mode CONACC --wig-scores --chrom $cName \
    -i SS $useGrp $ssSrc.ss > $file1.wigFix
popd > /dev/null
/bin/mkdir -p $out:h
sleep 4
/bin/touch $out:h
/bin/mv $tmp/$file1.wigFix $out
/bin/rm -fr $tmp
/bin/rmdir --ignore-fail-on-non-empty $cons/tmp/$grp/$d
/bin/rmdir --ignore-fail-on-non-empty $cons/tmp/$grp/$d:h
/bin/rmdir --ignore-fail-on-non-empty $cons/tmp/$grp
/bin/rmdir --ignore-fail-on-non-empty $cons/tmp
'_EOF_'
    # << happy emacs
    chmod +x doPhyloP.csh

    # Create list of chunks
    find ../../cons/SS/result -type f | grep ".ss$" \
	| sed -e "s/.ss$//; s#^../../cons/SS/result/##" > ss.list
    # make sure the list looks good
    wc -l ss.list
    #	19343 ss.list

    # Create template file
    #	file1 == $chr/$chunk/file name without .ss suffix
    cat << '_EOF_' > template
#LOOP
../run.phyloP/doPhyloP.csh $(path1) {check out line+ wigFix/$(dir1)/$(file1).wigFix}
#ENDLOOP
'_EOF_'
    # << happy emacs

    ######################   Running all species  #######################
    # setup run for all species
    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way/consPhyloP/all
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/consPhyloP/all
    rm -fr wigFix
    mkdir wigFix

    gensub2 ../run.phyloP/ss.list single ../run.phyloP/template jobList
    #	beware overwhelming the cluster with these fast running high I/O jobs
    para create jobList
    para try ... check ... push ... etc ...
    para -maxJob=53 push
    para time > run.time
# Completed: 19343 of 19343 jobs
# CPU time in finished jobs:     356438s    5940.63m    99.01h    4.13d  0.011 y
# IO & Wait Time:                129142s    2152.37m    35.87h    1.49d  0.004 y
# Average job time:                  25s       0.42m     0.01h    0.00d
# Longest finished job:             409s       6.82m     0.11h    0.00d
# Submission to last job:          8700s     145.00m     2.42h    0.10d

    mkdir downloads

    time (find ./wigFix -type f | sed -e "s#^./##; s#\.# d #g; s#-# m #;" \
	| sort -k1,1 -k3,3n | sed -e "s# d #.#g; s# m #-#g;" | xargs cat \
	| gzip -c > downloads/phyloP20way.wigFix.gz)
    #   real    39m3.271s

    # check integrity of data with wigToBigWig
    time (zcat downloads/phyloP20way.wigFix.gz \
	| wigToBigWig -verbose=2 stdin /hive/data/genomes/tarSyr2/chrom.sizes \
	phyloP20way.bw) > bigWig.log 2>&1
    egrep "real|VmPeak" bigWig.log
    # pid=53789: VmPeak:    22583408 kB
    # real    38m27.781s

    bigWigInfo phyloP20way.bw | sed -e 's/^/# /;'
# version: 4
# isCompressed: yes
# isSwapped: 0
# primaryDataSize: 4,465,722,302
# primaryIndexSize: 99,299,640
# zoomLevels: 10
# chromCount: 19343
# basesCovered: 2,059,732,828
# mean: -0.066487
# min: -15.955000
# max: 0.892000
# std: 0.737177

    #	encode those files into wiggle data
    time (zcat downloads/phyloP20way.wigFix.gz \
	| wigEncode stdin phyloP20way.wig phyloP20way.wib)
    # Converted stdin, upper limit 0.89, lower limit -15.96
    # real    11m39.490s

    du -hsc *.wi?
    # 2.0G    phyloP20way.wib
    # 319M    phyloP20way.wig

    # Load gbdb and database with wiggle.
    ln -s `pwd`/phyloP20way.wib /gbdb/tarSyr2/multiz20way/phyloP20way.wib
    time hgLoadWiggle -pathPrefix=/gbdb/tarSyr2/multiz20way tarSyr2 \
	phyloP20way phyloP20way.wig
    # real    0m36.219s

    # use to set trackDb.ra entries for wiggle min and max
    # and verify table is loaded correctly

    wigTableStats.sh tarSyr2 phyloP20way
# db.table      min max mean count sumData
# tarSyr2.phyloP20way     -15.955 0.892 -0.0664866 2059732828 -1.36945e+08
#       stdDev viewLimits
#     0.737177 viewLimits=-3.75237:0.892

    #	that range is: 15.955+0.892 = 16.847 for hBinSize=0.016847

    #  Create histogram to get an overview of all the data
    time hgWiggle -doHistogram \
	-hBinSize=0.016847 -hBinCount=1000 -hMinVal=-15.955 -verbose=2 \
	    -db=tarSyr2 phyloP20way > histogram.data 2>&1
    # real    6m48.616s

    # find the Y range for the 2:5 graph
    grep -v chrom histogram.data | grep "^[0-9]" | ave -col=5 stdin \
      | sed -e 's/^/# /;'
# Q1 0.000000
# median 0.000001
# Q3 0.000103
# average 0.001112
# min 0.000000
# max 0.025033
# count 899
# total 0.999981
# standard deviation 0.003201

    # find the X range for the 2:5 graph
    grep "^[0-9]" histogram.data | ave -col=2 stdin \
      | sed -e 's/^/# /;'
# Q1 -10.446000
# median -6.672300
# Q3 -2.864880
# average -6.691942
# min -15.955000
# max 0.892000
# count 899
# total -6016.056080
# standard deviation 4.408891

    #	create plot of histogram:
    cat << '_EOF_' | gnuplot > histo.png
set terminal png small x000000 xffffff xc000ff x66ff66 xffff00 x00ffff font \
"/usr/share/fonts/default/Type1/n022004l.pfb"
set size 1.4, 0.8
set key left box
set grid noxtics
set grid ytics
set title " Human tarSyr2 Histogram phyloP20way track"
set xlabel " phyloP20way score"
set ylabel " Relative Frequency"
set y2label " Cumulative Relative Frequency (CRF)"
set y2range [0:1]
set y2tics
set xtics
set xrange [-3:0.9]
set yrange [0:0.025]

plot "histogram.data" using 2:5 title " RelFreq" with impulses, \
        "histogram.data" using 2:7 axes x1y2 title " CRF" with lines
'_EOF_'
    #	<< happy emacs

    display histo.png &
    # appears to have an odd hole in the data near X=0 ?

#############################################################################
# construct download files for 20-way (DONE - 2015-04-15 - Hiram)
    mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/tarSyr2/multiz20way
    mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/tarSyr2/phastCons20way
    mkdir /usr/local/apache/htdocs-hgdownload/goldenPath/tarSyr2/phyloP20way
    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way/downloads
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/downloads
    mkdir multiz20way phastCons20way phyloP20way
    cd multiz20way
    time cp -p ../../anno/tarSyr2.20way.maf .
    #   real    2m22.776s
    # -rw-rw-r-- 1 84345995664 Apr 14 09:27 tarSyr2.20way.maf

    du -hsc *
    #  79G     tarSyr2.20way.maf

    time gzip *.maf
    #   real    87m37.865s
    # -rw-rw-r-- 1 10893506431 Apr 14 09:27 tarSyr2.20way.maf.gz

    grep TREE ../../4d/all.mod | awk '{print $NF}' \
      | ~/kent/src/hg/utils/phyloTrees/asciiTree.pl /dev/stdin \
         > tarSyr2.20way.nh
    ~/kent/src/hg/utils/phyloTrees/commonNames.sh tarSyr2.20way.nh \
      | ~/kent/src/hg/utils/phyloTrees/asciiTree.pl /dev/stdin \
         > tarSyr2.20way.commonNames.nh
    ~/kent/src/hg/utils/phyloTrees/scientificNames.sh tarSyr2.20way.nh \
	| $HOME/kent/src/hg/utils/phyloTrees/asciiTree.pl /dev/stdin \
	    > tarSyr2.20way.scientificNames.nh
    time md5sum *.nh *.maf.gz > md5sum.txt
    #   real    0m35.144s

    ln -s `pwd`/* \
        /usr/local/apache/htdocs-hgdownload/goldenPath/tarSyr2/multiz20way

    du -hsc *.maf.gz ../../anno/tarSyr2.20way.maf
    #  11G     tarSyr2.20way.maf.gz
    #  79G     ../../anno/tarSyr2.20way.maf

    # obtain the README.txt from tarSyr2/multiz17way and update for this
    #   situation

    #####################################################################
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/downloads/phastCons20way

    ln -s ../../cons/all/downloads/phastCons20way.wigFix.gz \
        ./tarSyr2.phastCons20way.wigFix.gz
    ln -s ../../cons/all/phastCons20way.bw ./tarSyr2.phastCons20way.bw
    ln -s ../../cons/all/all.mod ./tarSyr2.phastCons20way.mod
    time md5sum *.gz *.mod *.bw > md5sum.txt
    #   real    0m20.354s

    # obtain the README.txt from hg38/phastCons20way and update for this
    #   situation
    ln -s `pwd`/*.gz `pwd`/*.mod `pwd`/*.bw `pwd`/*.txt \
      /usr/local/apache/htdocs-hgdownload/goldenPath/tarSyr2/phastCons20way

    #####################################################################
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/downloads/phyloP20way

    ln -s ../../consPhyloP/all/downloads/phyloP20way.wigFix.gz \
        ./tarSyr2.phyloP20way.wigFix.gz
    ln -s ../../consPhyloP/run.phyloP/all.mod tarSyr2.phyloP20way.mod
    ln -s ../../consPhyloP/all/phyloP20way.bw tarSyr2.phyloP20way.bw

    time md5sum *.mod *.bw *.gz > md5sum.txt
    #   real    0m29.662s

    # obtain the README.txt from tarSyr2/phyloP17way and update for this
    #   situation
    ln -s `pwd`/* \
      /usr/local/apache/htdocs-hgdownload/goldenPath/tarSyr2/phyloP20way

    ###########################################################################
    ## create upstream refGene maf files
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/downloads/multiz20way
    # bash script
#!/bin/sh
export geneTbl="xenoRefGene"
for S in 1000 2000 5000
do
    echo "making upstream${S}.maf"
    featureBits tarSyr2 ${geneTbl}:upstream:${S} -fa=/dev/null -bed=stdout \
        | perl -wpe 's/_up[^\t]+/\t0/' | sort -k1,1 -k2,2n \
        | /cluster/bin/$MACHTYPE/mafFrags tarSyr2 multiz20way \
                stdin stdout \
                -orgs=/hive/data/genomes/tarSyr2/bed/multiz20way/species.list \
        | gzip -c > upstream${S}.${geneTbl}.maf.gz
    echo "done upstream${S}.${geneTbl}.maf.gz"
done
    #   real    82m21.917s

    md5sum upstream*.gz >> md5sum.txt

    # some other symlinks were already made above
    # obtain the README.txt from geoFor1/multiz20way and update for this
    #   situation
    ln -s `pwd`/upstream*.gz README.txt \
        /usr/local/apache/htdocs-hgdownload/goldenPath/tarSyr2/multiz20way

#############################################################################
# hgPal downloads (DONE - 2015-04-15 - Hiram)
#   FASTA from 20-way for knownGene, refGene and knownCanonical

    ssh hgwdev
    screen -S tarSyr2HgPal
    mkdir /hive/data/genomes/tarSyr2/bed/multiz20way/pal
    cd /hive/data/genomes/tarSyr2/bed/multiz20way/pal
    cat ../species.list | tr '[ ]' '[\n]' > order.list

    # this for loop takes about 2.5 hours on this large count contig assembly
    export mz=multiz20way
    export gp=xenoRefGene
    export db=tarSyr2
    export I=0
    export D=0
    mkdir exonAA exonNuc
    for C in `sort -nk2 ../../../chrom.sizes | cut -f1`
    do
        I=`echo $I | awk '{print $1+1}'`
        D=`echo $D | awk '{print $1+1}'`
        dNum=`echo $D | awk '{printf "%03d", int($1/1000)}'`
        mkdir -p exonNuc/${dNum} > /dev/null
        mkdir -p exonAA/${dNum} > /dev/null
	echo "mafGene -chrom=$C -exons -noTrans $db $mz $gp order.list stdout | gzip -c > exonNuc/${dNum}/$C.exonNuc.fa.gz &"
	echo "mafGene -chrom=$C -exons $db $mz $gp order.list stdout | gzip -c > exonAA/${dNum}/$C.exonAA.fa.gz &"
        if [ $I -gt 16 ]; then
            echo "date"
            echo "wait"
            I=0
        fi
    done > $gp.jobs
    echo "date" >> $gp.jobs
    echo "wait" >> $gp.jobs

    time sh -x ./$gp.jobs > $gp.jobs.log 2>&1 &
    # real    175m50.376s


    export mz=multiz20way
    export gp=xenoRefGene
    time find ./exonAA -type f | grep exonAA.fa.gz | xargs zcat \
     | gzip -c > $gp.$mz.exonAA.fa.gz
    # real    10m29.500s

    time find ./exonNuc -type f | grep exonNuc.fa.gz | xargs zcat \
     | gzip -c > $gp.$mz.exonNuc.fa.gz
    #   real    15m9.974s

  # -rw-rw-r--   1 611281544 Apr 15 20:37 xenoRefGene.multiz20way.exonAA.fa.gz
  # -rw-rw-r--   1 955571425 Apr 15 21:06 xenoRefGene.multiz20way.exonNuc.fa.gz

    export mz=multiz20way
    export gp=xenoRefGene
    export db=tarSyr2
    export pd=/usr/local/apache/htdocs-hgdownload/goldenPath/$db/$mz/alignments
    mkdir -p $pd
    md5sum *.fa.gz > md5sum.txt
    ln -s `pwd`/$gp.$mz.exonAA.fa.gz $pd/$gp.exonAA.fa.gz
    ln -s `pwd`/$gp.$mz.exonNuc.fa.gz $pd/$gp.exonNuc.fa.gz
    ln -s `pwd`/md5sum.txt $pd/

    rm -rf exonAA exonNuc

#############################################################################
# wiki page for 20-way (DONE - 2015-04-15 - Hiram)
    mkdir /hive/users/hiram/bigWays/tarSyr2.20way
    cd /hive/users/hiram/bigWays
    echo "tarSyr2" > tarSyr2.20way/ordered.list
    awk '{print $1}' /hive/data/genomes/tarSyr2/bed/multiz20way/20way.distances.txt \
       >> tarSyr2.20way/ordered.list

    # sizeStats.sh catches up the cached measurements required for data
    # in the tables.  They are usually already mostly done, only new
    # assemblies will have updates.
    ./sizeStats.sh tarSyr2.20way/ordered.list
    # dbDb.sh constructs tarSyr2.20way/TarSyr2_20-way_conservation_alignment.html
    # may need to add new assembly references to srcReference.list and
    # urlReference.list
    ./dbDb.sh tarSyr2 20way
    # sizeStats.pl constructs tarSyr2.20way/TarSyr2_20-way_Genome_size_statistics.html
    # this requires entries in coverage.list for new sequences
    ./sizeStats.pl tarSyr2 20way

    # defCheck.pl constructs TarSyr2_20-way_conservation_lastz_parameters.html
    ./defCheck.pl tarSyr2 20way

    # this constructs the html pages in tarSyr2.20way/:
# -rw-rw-r-- 1 10741 Apr 15 21:19 TarSyr2_20-way_conservation_alignment.html
# -rw-rw-r-- 1 14047 Apr 15 21:20 TarSyr2_20-way_Genome_size_statistics.html
# -rw-rw-r-- 1  8040 Apr 15 21:20 TarSyr2_20-way_conservation_lastz_parameters.html

    # add those pages to the genomewiki.  Their page names are the
    # names of the .html files without the .html:
#  TarSyr2_20-way_conservation_alignment
#  TarSyr2_20-way_Genome_size_statistics
#  TarSyr2_20-way_conservation_lastz_parameters

    # when you view the first one you enter, it will have links to the
    # missing two.

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