Figure 2
Fraction of correct fragment alignments as functions of the probabilities estimated from (2). For each main-chain fragment built, the sub-fragment with the highest weighted Z score was identified as described in §2. All alignments of this sub-fragment with the protein sequence were considered and the relative probabilities of each alignment were estimated with (2). An alignment was considered correct if the residue numbers of 90% of the residues in the fragment matched those of the nearest amino acid in the refined model. |