Homo Crystallographicus - Quo Vadis ?

Gerard J Kleywegt & T Alwyn Jones

Supplementary material to: G J Kleywegt & T A Jones (2002). Homo Crystallographicus - Quo Vadis ? Structure 10 (4), 465-472.

Raw data, statistics, and plots. The raw data was gathered and analysed using "jiffy" shell scripts and programs written by GJK. The plot files were generated with ODBMAN, and converted into PostScript with O2D.

New !!! Try Harry Plotter ... a Java-based interactive plotting program which provides direct links between data points in a scatter plot and the corresponding page at the RCSB-PDB !

1. master.list - ASCII text file containing the raw results of our analysis (~780 kB). The file contains one line per PDB entry with (tab-delimited): PDB identifier, resolution (Å), year of deposition, R-value, free R-value, number of amino-acid residues, number of Ramachandran-plot outliers, percentage of Ramachandran-plot outliers, method of test-set selection, number of test-set reflection, percentage of test-set reflections, (unused counter 1), (unused counter 2), flag to indicate presence or absence of electron-density map in EDS. The list is sorted by resolution first, and percentage Ramachandran-plot outliers second. The first few lines of this file look as follows:

    ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
   1ejg    0.54    2000    0.090   0.094   37      0       0.000   1_REFLECTION_OUT_OF_20  11220   5.000   10336   9119    HAS_AN_OMAP_IN_EDS
   1d8g    0.74    1999    0.105   0.131   0       0       0       _RANDOM_        NULL    NULL    10335   9118    NULL
   1dj4    0.75    1999    0.135   NULL    0       0       0       NULL    NULL    NULL    10333   9116    HAS_AN_OMAP_IN_EDS
    ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----

2. rval.out - output from the initial analysis program (~60 kB). This includes overall statistics, lists of PDB entries with the highest and lowest values for the various statistics, etc.

3. Analysis I - 10,674 entries with resolution and R-value.
   • rv.odb - O datablock file with raw data (~240 kB)
   • odbman_1.out output file from analysing this data with ODBMAN (~12 kB), containing histograms, correlation coefficient, and non-parametric rank correlations
   • Plot resolution_vs_year: PostScript | GIF
   • Plot rvalue_vs_year: PostScript | GIF
   • Plot rvalue_vs_resolution: PostScript | GIF
   • Plot hasmap_vs_year: PostScript | GIF

4. Analysis II - 6,560 entries with resolution, R-value and free R-value.
   • rvrf.odb - O datablock file with raw data (~250 kB)
   • odbman_2.out output file from analysing this data with ODBMAN (~18 kB), containing histograms, correlation coefficient, and non-parametric rank correlations
   • Plot rfree_vs_year: PostScript | GIF
   • Plot rfminusr_vs_year: PostScript | GIF
   • Plot rfree_vs_rvalue: PostScript | GIF
   • Plot rfminusr_vs_rvalue: PostScript | GIF
   • Plot rfminusr_vs_rfree: PostScript | GIF
   • Plot rfree_vs_resolution: PostScript | GIF
   • Plot rfminusr_vs_resolution: PostScript | GIF
   • Plot hasmap_vs_rfree: PostScript | GIF

5. Analysis III - 10,215 entries with resolution, R-value and Ramachandran plot.
   • rvra.odb - O datablock file with raw data (~475 kB)
   • odbman_3.out output file from analysing this data with ODBMAN (~21 kB), containing histograms, correlation coefficient, and non-parametric rank correlations
   • Plot percramaout_vs_year: PostScript | GIF
   • Plot percramaout_vs_resolution: PostScript | GIF
   • Plot percramaout_vs_rvalue: PostScript | GIF
   • Plot hasmap_vs_percramaout: PostScript | GIF
   • Plot numres_vs_year: PostScript | GIF

6. Analysis IV - 6,316 entries with resolution, R-value, free R-value, and Ramachandran plot.
   • rvrfra.odb - O datablock file with raw data (~390 kB)
   • odbman_4.out output file from analysing this data with ODBMAN (~27 kB), containing histograms, correlation coefficient, and non-parametric rank correlations
   • Plot percramaout_vs_rfree: PostScript | GIF
   • Plot percramaout_vs_rfminusr: PostScript | GIF

7. Analysis V - 5,421 entries with more than one protein or peptide chain (at least 10 residues).
   • anaram_max.log, list of entries with more than one protein or peptide chain, sorted by the percentage Ramachandran outliers of the "worst" chain (~545 kB), or just the Top 100. Entries look as shown below and list: index (ignore), PDB identifier, "best" chain with name, percentage outliers and number of residues, "worst" chain with name, percentage outliers and number of residues, the difference between the two percentages, the resolution (Å), and the year of deposition:

      ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
      #   1594 1f83 MIN   7.60 A   397 res - MAX  76.20 B    21 res - DIFF  68.60 - RESO   2.00 - YEAR 2000
      #   5389 1tmf MIN  27.00 2   244 res - MAX  56.00 4    25 res - DIFF  29.00 - RESO   3.50 - YEAR 1992
      #   5221 1nro MIN  27.60 H   225 res - MAX  50.00 L    22 res - DIFF  22.40 - RESO   3.10 - YEAR 1994
      #   5373 1fll MIN  18.70 A   187 res - MAX  50.00 X    18 res - DIFF  31.30 - RESO   3.50 - YEAR 2000
      #   4977 1awh MIN   7.90 B   229 res - MAX  44.40 A    27 res - DIFF  36.50 - RESO   3.00 - YEAR 1997
      ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----

   • anaram_max.log, list of entries with more than one protein or peptide chain, sorted by the difference between the percentage Ramachandran outliers of the "worst" and the "best" chain (~545 kB), or just the Top 100. A histogram of the differences looks as follows:

      ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
      Nr >=       0.0000 and <       1.0000 :     2152 ( 39.70 %; Cum  39.70 %)
      Nr >=       1.0000 and <       2.0000 :     1240 ( 22.87 %; Cum  62.57 %)
      Nr >=       2.0000 and <       3.0000 :      697 ( 12.86 %; Cum  75.43 %)
      Nr >=       3.0000 and <       4.0000 :      406 (  7.49 %; Cum  82.92 %)
      Nr >=       4.0000 and <       5.0000 :      243 (  4.48 %; Cum  87.40 %)
      Nr >=       5.0000 and <      10.0000 :      454 (  8.37 %; Cum  95.78 %)
      Nr >=      10.0000 and <      20.0000 :      172 (  3.17 %; Cum  98.95 %)
      Nr >=      20.0000 and <      30.0000 :       46 (  0.85 %; Cum  99.80 %)
      Nr >=      30.0000 and <      40.0000 :       10 (  0.18 %; Cum  99.98 %)
      Nr >=      60.0000 and <      70.0000 :        1 (  0.02 %; Cum 100.00 %)
      Nr >=      70.0000                    :        0 (  0.00 %; Cum 100.00 %)
      ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----

   • Plot percentage outliers "worst" chain versus "best" chain: PostScript | GIF
   • Plot difference percentage outliers "worst" chain and "best" chain versus resolution: PostScript | GIF