CPDB 'Rats and Mice' rows: 1534 | positive carcinogens (numeric TD50): 785
TD50 (mg/kg/day): median=19.3  geomean=16.5  min=2.35e-05  max=3.03e+04

--- Reproduction of Rulis's curve 5 (fraction of carcinogens with lifetime risk > 1e-6) ---
  at 5 ppb (a level 'suggested by some')           :  72.7% exceed 1e-6   (excluded: 27.3%)
  at 1 ppb (Rulis: ~half on either side)           :  52.7% exceed 1e-6   (excluded: 47.3%)
  at 0.5 ppb (the 1995 ToR rule)                   :  45.4% exceed 1e-6   (excluded: 54.6%)
  at 0.05 ppb = 50 ppt diet (Rulis's worked example):  20.3% exceed 1e-6   (excluded: 79.7%)
  level holding the least-potent 50% of carcinogens under 1e-6: 0.772 ppb (the most-potent 50% would exceed)
  level holding the least-potent 85% of carcinogens under 1e-6: 0.0293 ppb (the most-potent 15% would exceed)
  level holding the least-potent 90% of carcinogens under 1e-6: 0.0148 ppb (the most-potent 10% would exceed)
  level holding the least-potent 95% of carcinogens under 1e-6: 0.00386 ppb (the most-potent 5% would exceed)

  Rulis bookkeeping at 0.5 ppb: if 1 in 5 unknown migrants is a carcinogen, share of ToR decisions >1e-6 = 0.2 x 0.45 = 9.1%  (=> ~91 of 100 decisions <= 1e-6)

Frawley appendix: parsed ~41 distinct compound names (of 220)
Frawley compounds that are CPDB carcinogens — exact-name matches: 2
   = d c red no
   = toxaphene
...additional likely matches (substring): 16
   ~ ponceau sx no -> s
   ~ polyoxyethylene stearate a -> s
   ~ citrus red no no -> s
   ~ sodmm alkylbenzenesulphonatc -> benzene
   ~ selenium -> selenium sulfide
   ~ sodium nitratc -> s
   ~ phosphorothioate -> s
   ~ terpene polychlorinates -> s
   ~ wood rosin dark -> s
   ~ sorbitan monostearate -> s
   ~ polyoxyethylene stearate -> s
   ~ ethanols triton x 405p -> s
   ~ sorbitan monopalmitate -> s
   ~ sulphonate -> s
   ~ ammonium sulphamate -> s
   ~ monostearate -> s