A five-step annotation methodology was used.
Step 1: New sequence was obtained through pyrosequencing. Step 2: Using a word-counting strategy, the best-match V was determined. When there was a tie for best match, or when the best match was nevertheless a poor match, sequences were removed from subsequent V-D-J usage analysis. The best-match V was then aligned to the new sequence using megablast. Steps 3: The region that matched V was removed (grayed out in the figure) and the same word-counting strategy was applied to the remaining sequence to determine the best-match J sequence. As with V, ties and poor matches were removed; the best-match J was aligned using megablast. Step 4: The region that matched J was removed (gray) and D was assigned and aligned to the remainder as in Steps 2 and 3. Step 5: Using the alignments, exonucleotide chewback and nucleotide addition were determined.