Skip to main content
. Author manuscript; available in PMC: 2009 Aug 1.
Published in final edited form as: J Proteome Res. 2008 Jun 19;7(8):3373–3381. doi: 10.1021/pr800129m

Table 2.

Conditional Probability Table a

PCID(Ij|Iπ(j), S, Pj)
Iπ(j) intensity Ij intensity Ij = b − H3PO4
Iπ(j) = b
Ij = y − H3PO4
Iπ(j) = y
strong strong 22.5% 9.5%
strong medium 20.4% 15.2%
strong weak 3.8% 4.8%
strong absent 53.3% 70.5%
medium strong 8.1% 1.7%
medium medium 25.1% 11.4%
medium weak 10.8% 8.6%
medium absent 55.9% 78.4%
weak strong 3.1% 0.4%
weak medium 12.9% 4.0%
weak weak 14.5% 9.4%
weak absent 69.5% 86.3%
absent strong 5.4% 1.5%
absent medium 15.9% 4.1%
absent weak 9.3% 4.4%
absent absent 69.4% 90.0%
a

Using the observed intensity for Ij and Iπ(j), we look up the learned conditional probability and score the peak assignment Ij accordingly. This table shows two conditional probabilities: the third column for b − H3PO4 given b, the fourth column for y − H3PO4 given y. Notice the distinct propensities for fragment neutral loss of a b ion compared to a y ion. For example, a strong b peak produces a medium/strong b − H3PO4 43% of the time. Remembering that on average only 50% of b ions contain the phosphate moiety, almost all phosphorylated b peaks are accompanied by a neutral phosphate loss. In contrast, a strong y peak produces a medium/strong y − H3PO4 only 25% of the time, or roughly half of the fragments containing a phosphate.