Successful de novo sequencing algorithms are more diverse than database searching algorithms. A common initial step is to reduce the experimental mass list size by trying to remove noisy peaks. Additional preprocessing steps can be considered like normalizing intensities or trying to improve experimental peptide mass accuracy . As a second step, there are algorithms that try to predict the ion type corresponding to each experimental mass . As a matter of fact, if we know which masses in a spectrum are b or y ions, then we can read the sequence directly (see Figure 9.4). In practice, this step cannot be perfect, and numerous additional masses are predicted as b or y and hence the problem difficulty is reduced only.
After this optional second step, four main types of methods are employed. The most common is the spectrum graph: A graph is built by creating one node per experimental mass plus one for the entire peptide and one empty node. The nodes, whose masses differ by one amino acid mass, are linked and the edges labeled with the amino acids (see Figure 9.4). In the graph, the correct peptide sequence is assumed to appear as a path from the empty node to the entire peptide node. Due to possibly missing b or y fragment masses, which would disrupt the correct path, experimental masses are complemented before the graph is built: m ^ mass(P) + 2 - m, where mass(P) the mass of the peptide P. This complementation changes the mass of a b ion into the complementary y ion and vice versa. The spectrum graph is then exploited either empirically or—more frequently—by scoring each edge and by searching for the longest path . To score edges requires scoring functions similar to the
Was this article helpful?
Do you hate the spring? Do you run at the site of a dog or cat? Do you carry around tissues wherever you go? Youre not alone. 51 Ways to Reduce Allergies can help. Find all these tips and more Start putting those tissues away. Get Your Copy Of 51 Ways to Reduce Allergies Today.