Weighted CSI program

written by Kyoko Yap

Explanation

CSI = 0.25*(1*dCaCb[i-1]+2*dCaCb[i]+1*dCaCb[i+1])

where

dCaCb[i] = dCa[i] - dCb[i]

and

dCa[i] = secondary chemical shift
= carbon-alpha chemical shift for residue i, minus carbon-alpha average random coil chemical shift for that residue type

Appropriate weighting if values are missing. If chemical shifts for Ca or Cb are missing for residue i, the CSI for that residue is zero. If they are missing for neighbouring (i-1 or i+1) residues, averaging for CSI calculation is adjusted accordingly.

Download

for one-letter amino acid names (calc_csi)
Sun/Solaris binary
SGI binary
Linux binary

for three-letter amino acid names (calc_csi_long)
Sun/Solaris binary
SGI binary
Linux binary

for BMRB input (bmrb2csi)

Example input

amino acid / residue number / carbon-alpha shift / carbon-beta shift

for calc_csi

A      1    51.9    19.4
D      2    54.7    41.5
Q      3    55.6   29.9
L      4    54.6    43.7
T      5    60.5    71.1
E      6    59.9    29.3

for calc_csi_long

ALA      1    51.9    19.4
ASP       2    54.7    41.5
GLN      3    55.6   29.9
LEU      4    54.6    43.7
THR      5    60.5    71.1
GLU      6    59.9    29.3

Executing the program

kyap[35] calc_csi.sol cam.cs r > cam.csi

Use "r" if your cysteines are reduced (i.e. Cb is around 29, not 41), otherwise use "o".

Important notes

Missing values and Cb values for Gly (i.e. zero) should be listed as zero. The program needs values in all columns, zero or otherwise.

Missing residues need to be in the input file, or the calculated CSI value may be wrong. If you do not know the amino acid identity of the missing residues, 'X' is fine:

A      1    51.9    19.4
D      2    54.7    41.5
X      3    0        0
X      4    0        0
X      5    0        0
E      6    59.9    29.3

Please e-mail Kyoko if you have problems.

Weighted CSI program originally based on code by Chieri Tomomori.