Thursday, July 21, 2016

Finding disordered residues in an NMR ensemble

Note to self: here's how you identified disordered residues in the NMR ensemble 2KCU.pdb

1. In Pymol: "fetch 2kcu"

2. Action > align > states (*/CA)
2016.08.07 update: the above command also aligns the tails.  Use "intra_fit (2kzn///6-158/CA)"

3. "save 2kcu_aligned.pdb, state=0"

4. In terminal: grep CA 2kcu_aligned.pdb > lis

5. python disorder.py

disorder.py (given below) calculates the standard deviation of the x, y, and z coordinate of each CA atom ($\sigma_{x,i}, \sigma_{y,i}, \sigma_{z,i})$. It then averages these three standard deviations for each CA atom $(\sigma_i)$.  To find outliers, it averages these values for the entire protein $(\langle \sigma_i \rangle)$ and computes the standard deviation of this average $(\sigma_{\langle \sigma_i \rangle})$. Any residues for which $\sigma_i > \langle \sigma_i \rangle + \sigma_{\langle \sigma_i \rangle}$ is identified as disordered.

Here I've colored the disordered residues red (haven't updated the picture based on Step 2-change yet)



Yes, I know: "the 1970's called and want their Fortran code back". How very droll.



This work is licensed under a Creative Commons Attribution 4.0

No comments: