# How can the Autocovariances, autocorrelations, and autocorrelation coefficients be calculated from a Protein Amino Acid Sequence?

Given a normal protein sequence with the 20 standard amino acids, how can the 'Autocovariances', 'autocorrelations', and 'autocorrelation coefficients' of the sequence be calculated?

What is meant by these terms in the context of a protein sequence?

What are the appropriate algorithms and formulas?

Is there any existing package/function to calculate these values?

• Could you provide some context? It could clarify the question and attract more answers. Maybe you read that in an article or it is for some specific goal in mind, or someone asked you this because a discussion about another topic
– llrs
Sep 24 '18 at 14:28

Protein autocorrelation is similar to usual autocorrelation of time series (that is, function $$r(k)$$ measuring correlation between values lagged by $$k$$), except that instead of time dimension we slide along the backbone of amino acid chain. The "dependent variable" can be any quantifiable property of the amino acids - then the autocorrelation measures similarity in that particular property.