The original question: If we have a sequence of real numbers and on that, we define a sequence of $n^{th}$ rational numbers and another of irrational numbers, then they are both subsequences of the original sequence of real numbers?
In other words, can a rational sequence be a complementary subsequence of an irrational sequence?
That is can there exist a sequence of real numbers which has both a rational subsequence and an irrational subsequence?
Yes. Take the sequence of irrational numbers $\frac{1}{\sqrt{p}}$ as $p\rightarrow \infty$, which converges to zero. Now consider the sequence of rational numbers $\frac{1}{p}$, as $p\rightarrow \infty$, which converges to zero.
Sure. When constructing sequences, you have great freedom. As an exercise, do a $\delta - \epsilon$ proof that the sequence ${\frac{1}{p_1}, \frac{1}{\sqrt{p_1}}, \frac{1}{p_2}, \frac{1}{\sqrt{p_2}},...}$ converges to zero. Then show each "compliment" sub-sequence converges to zero.
Here's another thing you could maybe think about playing with, the AM-GM inequality: $$\frac{a_1 + a_2 + a_3 + \cdots + a_n}{n} \geq \sqrt[n]{a_1\cdot a_2 \cdot a_3 \cdot \cdots \cdot a_n} $$