the sequence A006743

[Marc LeBrun]

 >=Thomas Baruchel
 > Have you ever noticed that the  coefficients of the linear
 > recurrence are given by the coefficients in the denominator

My apologies if I'm just restating the obvious (I'm not very adept with 
Pade approximations), but the generating function for any linear recurrence 
is always a rational function whose denominator coefficients are those of 
the recurrence and whose numerator coefficients reflect the initial values, 
right?