This tool analyzes tolerance to substitutions at all positions of a protein sequence, searching for correlations with amino acid descriptors. For a given dataset (from alignments, deep-sequencing experiments, etc.) the main tool will find out which property(ies) of the amino acids (volume, hydrophobicity, etc.) shape the probability of finding each amino acid at each position of the sequence. Such kind of analysis helps retrieve rich information about the main physicochemical constraints to protein design, engineering and evolution, as illustrated in this application to deep-sequencing data on a library of all mutants at all positions of an enzyme.

The main data is a table of:

ΔΔG_aa,i = -log(P_aa,i/P_wt,i)

where P_aa,i = fractional occurrence of amino acid type aa (alphabetically ordered from A=1 to Y=20) at site i in the protein sequence and P_wt,i = fractional occurrence of the 'wild type' (or reference) amino acid.

This tool was originally developed to dissect protein variability data from deep-sequencing experiments as done in this work but is also helpful for understanding variability in alignments. This other tool converts alignments to ΔΔG data.

What's the meaning of the picture on the home page?

The main page image illustrates how protein traits emerge from amino acid properties.