journal menu
![[Figure 6]](yn5112fig6mag.jpg)
|
|
Figure 6
Benchmarking the effectiveness of compression strategies for HIV reverse transcriptase using binary metrics. The software package PHENIX (Adams et al., 2010  ) was used to compare the electron density obtained from uncompressed data with the electron density obtained from data that were compressed using various strategies. The binary metrics (i) poor fit and (ii) conformational flip were added to indicate the total number of amino acids that were flagged as incorrect (hence the Y axis shows the number of amino acids that did not fit the observed electron density plus the number of amino acids that only fit with a different conformation). Just as in the case of Fig. 5, the slope of the decrease in quality varies between different compression strategies (and a smaller slope indicates a better compression strategy). Significantly, the best overall performance (labelled BIN2_SUM2_HCOMP) was the same when benchmarked using real space R values and when benchmarked using binary metrics. |
![[Figure 6]](yn5112fig6.jpg)
 | JOURNAL OF SYNCHROTRON RADIATION |
ISSN: 1600-5775
access
