search

lindsly / 4DNvestigator

MIT License
0 stars 0 forks source link

How to calculate the Von Neumann Entropy for a given region (or several given regions) of the genome #3

Open kuangzhuoran opened 1 year ago

kuangzhuoran commented 1 year ago

I want to compare VNE (Von Neumann Entropy) in different regions on the genome.

In both entropyExample.m and entropyExampleExpanded, VNE can only be calculated for one chromosome I note that in vne.m. tempEig = eig(A); tempEig = tempEig/(sum(tempEig)); vneOut = -sum(tempEig.*log(tempEig));

I ended up looking through your published paper "4DNvestigator: time series genomic data analysis toolbox“ In the section "Network entropy", the formula for Entropy is given Considering the contents of vne.m, so I only need to calculate eigenvalues for the specified region: a, VNE for the specified region = - a * ln(a) The question is, what if a < 0？

kuangzhuoran commented 1 year ago

there are a number of regions that I am interested in, and they are all special regions that I have filtered out, so each region calculates a value and then just sums it up. This is also consistent with the article, vne.m

"tempEig = tempEig/(sum(tempEig))" Another question is, if I want to calculate the VNE for several given regions, is sum(tempEig) here a genome-wide calculation of the sum of eigenvalues

kuangzhuoran commented 1 year ago

Or I can just calculate the correlation matrix for the given regions, use this matrix to calculate the eigenvalues, and then calculate the VNE