The scaSectorID script does the preliminaries of sector identification and stores the outputs using the python tool pickle:

  1. Chooses \(k_{max}\) (the number of significant eigenmodes) by comparison of the \(\tilde{C_{ij}}\) eigenspectrum to that for the randomized matrices

  2. Rotates the top \(k_{max}\) eigenvectors using independent components analysis

  3. Defines the amino acid positions that significantly contribute to each of the independent components (ICs) by empirically fitting each IC to the t-distribution and selecting positions with greater than a specified cutoff (default: p=0.95) on the CDF.

  4. Assign positions into groups based on the independent component with which it has the greatest degree of co-evolution.

Key Arguments
--input, -i

*.db (the database produced by running scaCore)

--kpos, -k

number of significant eigenmodes for analysis (the default is to automatically choose using the eigenspectrum)

--cutoff, -p

empirically chosen cutoff for selecting AA positions with a significant contribution to each IC, Default = 0.95

--matlab, -m

write out the results of this script to a matlab workspace for further analysis


scaSectorID -i PF00071_full.db

Kim Reynolds



Copyright (C) 2015 Olivier Rivoire, Rama Ranganathan, Kimberly Reynolds

This program is free software distributed under the BSD 3-clause license, please see the file LICENSE for details.