The phrase "accessibility class" is correct and usable in written English.
It can be used in contexts related to programming, web development, or education, where it refers to a classification that indicates how accessible a particular element or feature is for users with disabilities.
Example: "In our web application, we implemented an accessibility class to ensure that all users can navigate the site effectively, regardless of their abilities."
Alternatives: "accessibility category" or "accessibility level".
Indeed, the μ a terms can be seen as an additive constant to each accessibility term for a given accessibility class (see additional file 6).
For each model, we computed the accessibility class of each residue, as was done for the structures in the training set (see Section 2.1).
The second term represents the accessibility free energy: ν i is the accessibility class of the site i and is the solvent accessibility potential of the amino acid a when placed into the accessibility class d (d = {1.. D}), where D is the number of accessibility classes.
The information in this dataset was used to calculate a propensity matrix, giving the compatibility of each conservation class with each accessibility class (Section 2.3, and Supplementary Table S1).
Concerning the accessibility potentials, it is important to note that we are working in a protein design context (i.e. we are evaluating the fitness of alternatives amino acids in a given accessibility class).
In addition, our score is based on summation of information that is local in the structure (the propensity of the conservation class of each residue for its accessibility class), so it should be adaptable to provide a local quality score for each residue of the structure, as is done by local quality assessment tools (Fasnacht et al., 2007).
Each residue was then classified into one of ten evenly distributed accessibility classes.
To determine the accessibility classes, we first compute the solvent accessibility area using Naccess 2.1 [ 35] and partitioned the resulting values into classes [ 26].
Figures 3 6 show breakdowns of the degree of inconsistency for functional residues, residue classes, solvent accessibility classes and distance from gaps (see Section 2).
ConQuass assigns each structure the average score of its residues: where C and A are vectors of the same length L (number of amino acids in the protein), giving, respectively, the conservation and accessibility classes of the residues.
If one wants to compare the accessibility potentials between classes for a given amino acid (i.e. in a protein folding perspective), one solution is to remove the logarithm of the frequency of the accessibility classes to each potential (additional file 4).
