Fig. 9. Principle of one iteration of the stack algorithm for DNA: an input node I is extended for each of the possible bases (A, C, G, and T) at the current position.

The new decoding branches (I+A, I+G, I+C, and I+T) are then inserted into the decoding stack, and the stack is sorted according to the branch metrics. The node(s) with the highest metric in the stack will be extended in the next iteration.