=item Daphnia pulex tests   2007 june 16, tandy v.5q

  melon:/bio/bio-grid/mb/EVidenceModeler/daphd

  scaffold_1 size=4193030   tandy HSP=876  ; match=202  ; region=114
  scaffold_2 size=3740169   tandy HSP=1039 ; match=300  ; region=188
  scaffold_3 size=3777634   tandy HSP=1255 ; match=320  ; region=170
  scaffold_4 size=3075709   tandy HSP=1894 ; match=501  ; region=249
  scaffold_5 size=2511979   tandy HSP=1012 ; match=180  ; region=81
  scaffold_6 size=2406117   tandy HSP=1576 ; match=390  ; region=189
  scaffold_7 size=2324446   tandy HSP=1016 ; match=260  ; region=124
  scaffold_8 size=2335496   tandy HSP=705  ; match=167  ; region=92
  scaffold_9 size=2251199   tandy HSP=1345 ; match=389  ; region=222
  26 MB total

  Gene counts
    cat scaffold_?/dpulex1_predict.gff | grep mRNA | perl -ne'($r,$s,$t)=split; print "$s\n" if($t eq "mRNA");' | sort | uniq -c
    11196 DGIL_SNO  << this is 2x data stutter;  5598 is right
    3315 JGI
    4332 NCBI_GNO
    2381 twinscan   ?? drop these
  use 4332 as median gene count

  # genebest for Dpulex scaffold 1-9
  # gene bestids
  scores/method  (num is all ids x all gene matches, ~ 2/match)
  predictors      num     score   lost    tandem  near    equal   
  DP_DGIL_SNO_    4394    64.36   6.89    0.60    0.20    0.31
  Dappu           1821    63.55   2.56    0.58    0.24    0.31
  NCBI_GNO_       3333    63.38   2.88    0.77    0.24    0.28
  
  # geneinfo for Dpulex scaffold 1-9
  # gene info, all-methods
  scores/method   (count is all gene matches)
  predictor       count   eqfar   eqnear  eqsame  fullgns gnids  maxgns  methods nexons  tandkno tandnew  tandems
  DP_DGIL_SNO_    2296    1.00    1.22    3.32    1.51    3.89    2.11    1.81    4.33    371     103     0.44
  Dappu           945     1.31    2.12    5.75    3.17    6.90    2.99    2.71    6.85    286     48      0.73
  NCBI_GNO_       1411    1.14    1.74    4.56    2.33    5.56    2.65    2.34    5.63    331     86      0.64
  all             2709    0.96    1.11    2.95    1.34    3.52    1.99    1.72    3.96    383     121     0.42

  # .............................................................
  Using 501 for all known/new tandems, and 4332 gene predictions for these 9 scaffolds gives tandem rate of ~ 12%
    (is this calc right? known tandems are included in base count of 4000, but not 'new' tandem models)