I am working with Scoring matrices and its evaluation. Its very important to understand the concept of entropy values and how it is derived. I know that, how the entropies were calculated for Blosum matrices, using background frequency and score for each amino acid pair. But I am puzzled how it has been calculated in case of PAM matrix. The original paper by Dayhoff have not involved in calculating this values, and carries no information regarding this. But I do find the entropy values for each matrix PAM available in NCBI ftpsite. ftp://ftp.ncbi.nlm.nih.gov/blast/matrices/PAM250
Could anyone please explain how it is derived in PAM.
Dear Reka,
see this source, pages (6,7,8, 9, 10)
Most of the slides/pages have references to original articles at the bottom of these pages
https://www.cs.umd.edu/class/fall2011/cmsc858s/Substitution_Matrices.pdf
how to generate PAM-marrix:
Where Is Pam 1.0.7 Source (To Generate Pam Substitution Matrices)?
this post also discusses many marices
A: Statistical Comparison Between Position-Specific Scoring Matrices
Comparison of Blosum and PAM-calculation:
http://www.cs.columbia.edu/4761/assignments/assignment1/reference1.pdf
If this is not enough, see similar posts on the right panel.
Thanks for the reply. I have referred to this material previously and its clear it has been calculated using back ground frequency and amino acid score(Sij). In case of PAM, I am puzzled with the Sij, value. Is it the value, derived before the scaling factor?