Sequences with divisibility A sequence is such that $a_o =1, a_1= 1, a_{n+1}
=a_{n}a_{n-1}+1$ so we have to comment on divisibilty of $a_{2007} $ by 4.
I found out first few values in sequence as 
1,1,2,3,7,22, .... which told me that only $a_{3n} $ is even.
But can there be some other elaborative method?
 A: The sequence $a_n\bmod 4$ is eventually periodic with a readily determined period ...
A: Hint $ $ There are $j<k$ with $\,(a_j,a_{j+1})\equiv (a_k,a_{k+1})\,\pmod{\!4}\ $ since there are only finitely many such pairs $\!\bmod {\!4}$. The recurrence depends only on the pair of prior values so this leads to cyclic behavior. 
A: Questions regarding divisibility in general are often a lot easier to tackle using modulaic algebra since we don't care about precisely what values the numbers will assume. In case you don't know what the modulo k operator is, it is basically the remainder that you get after you divide a number by k. If a number is divisible by k, so in our case 4, then it is zero in mod 4 since there is no rest. 
So in this case we can translate the first values that you got into mod 4 as follows: 1,1,2,3,3,2 ... Now let's multiply a number in mod 2 by a number in mod 3. These numbers can be written as a=(4c+2) and b=(4d+3), so the result of their multiplication is 4d+6. However, we should add an extra one in the sequence, which will give us 4d+7. As you can verify, this number is indeed 3 in mod 4. Whenever we multiply two numbers that are 3 in mod 4, the next number in the sequence will be 2. 
Therefore we can conjecture that no element in the entire sequence will be divisible by 4 since each one of them will be either 2 or 3 in mod 4.
