Department of Cellular and Molecular Pharmacology
University of California, San Francisco
Acta Crystallographica Section D 55: 506-517
WWW Service:
WebMol (Option 'Rama', map type: 'df/dRes')
Supplementary material:
gif-format:
Ala,
Cys,
Asp,
Glu,
Phe,
Gly,
His,
Ile,
Lys,
Leu,
Met,
Asn,
Pro,
Gln,
Arg,
Ser,
Thr,
Val,
Trp,
Tyr
postscript format:
Ala,
Cys,
Asp,
Glu,
Phe,
Gly,
His,
Ile,
Lys,
Leu,
Met,
Asn,
Pro,
Gln,
Arg,
Ser,
Thr,
Val,
Trp,
Tyr
Note/ Correction:
In the concluding remarks of chapter
3.1 we state that "the fractional
content of any secondary structural
type defined in DSSP was observed to
be uncorrelated with resolution
with corresponding correlation
coefficients of r<0.03 for each
secondary structural type including
random coil interpreted as a separate
structural category. Thus, the
frequency shifts observed in Figure
1B did not originate from an
unbalanced data set."
We note that the reported maximal
absolute value of 0.03 actually was
obtained for the regression slopes
and not the correlation coefficients
r.
For the correlation coefficients we
find |r|<0.05 for all secondary
structural states defined in DSSP
except for r=-0.19 for Turns (designated 'T') and
r=-0.11 for isolated beta-bridges (designated 'B').
The conclusion in our original paragraph remains
unchanged. However, the significant
correlation coefficient for Turns
(increased frequencies at higher resolution)
shows the tendency towards
all satisfied hydrogen bonds in high-resolution
protein structures.