How much correlation can we expect?

Use r values from closely related species to estimate 
'good' values, and from totally unrelated species to 
estimate 'bad' values.

human       chimp    0.8479
human       mouse    0.9929 <- ?
human       tomato   0.3424
chimp       tobacco  0.0099
tobacco     tomato   0.9889

There is probably significant sampling error, fewer chimp
sequences available than mouse, human, or tobacco. 
The numbers are not exactly what we'd expect, but can 
provide a rough guide.