Our algorithm finds the closest matches based on actives, their concentration and the full formula similarity
