aa20.hpp

Go to the documentation of this file.

// -----------------------------------------------------------------------------------------------------
// Copyright (c) 2022 deCODE Genetics
// Copyright (c) 2006-2020, Knut Reinert & Freie Universität Berlin
// Copyright (c) 2016-2020, Knut Reinert & MPI für molekulare Genetik
// This file may be used, modified and/or redistributed under the terms of the 3-clause BSD-License
// shipped with this file and also available at: https://github.com/biocpp/biocpp-core/blob/main/LICENSE.md
// -----------------------------------------------------------------------------------------------------
 
#pragma once
 
#include <vector>
 
#include <bio/alphabet/aminoacid/aminoacid_base.hpp>
#include <bio/alphabet/aminoacid/concept.hpp>
#include <bio/alphabet/detail/to_lower.hpp>
 
namespace bio::alphabet
{
 
class aa20 : public aminoacid_base<aa20, 20>
{
private:
    using base_t = aminoacid_base<aa20, 20>;
 
    friend base_t;
    friend base_t::base_t;
 
public:
    constexpr aa20() noexcept                         = default; 
    constexpr aa20(aa20 const &) noexcept             = default; 
    constexpr aa20(aa20 &&) noexcept                  = default; 
    constexpr aa20 & operator=(aa20 const &) noexcept = default; 
    constexpr aa20 & operator=(aa20 &&) noexcept      = default; 
    ~aa20() noexcept                                  = default; 
 
    using base_t::base_t;
 
protected:
    static constexpr std::array<char_type, alphabet_size> rank_to_char{
      'A', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'K', 'L', 'M', 'N', 'P', 'Q', 'R', 'S', 'T', 'V', 'W', 'Y',
    };
 
    static constexpr std::array<rank_type, 256> char_to_rank = []() constexpr
    {
        std::array<rank_type, 256> ret{};
 
        // initialize with UNKNOWN
        ret.fill(15); // value of 'S', because that appears most frequently
 
        // reverse mapping for characters and their lowercase
        for (rank_type rnk = 0u; rnk < alphabet_size; ++rnk)
        {
            ret[static_cast<rank_type>(rank_to_char[rnk])]                   = rnk;
            ret[static_cast<rank_type>(detail::to_lower(rank_to_char[rnk]))] = rnk;
        }
 
        ret['B'] = ret['D'];
        ret['b'] = ret['D']; // Convert b (either D/N) to D, since D occurs more frequently.
        ret['J'] = ret['L'];
        ret['j'] = ret['L']; // Convert j (either I/L) to L, since L occurs more frequently.
        ret['O'] = ret['L'];
        ret['o'] = ret['L']; // Convert Pyrrolysine to lysine.
        ret['U'] = ret['C'];
        ret['u'] = ret['C']; // Convert Selenocysteine to cysteine.
        ret['X'] = ret['S'];
        ret['x'] = ret['S']; // Convert unknown amino acids to serine.
        ret['Z'] = ret['E'];
        ret['z'] = ret['E']; // Convert z (either E/Q) to E, since E occurs more frequently.
        ret['*'] = ret['W']; // The most common stop codon is UGA. This is most similar to a Tryptophan.
        return ret;
    }();
};
 
} // namespace bio::alphabet
 
// ------------------------------------------------------------------
// literals
// ------------------------------------------------------------------
 
namespace bio::alphabet
{
 
inline namespace literals
{
 
consteval aa20 operator""_aa20(char const c)
{
    if (!char_is_valid_for<aa20>(c))
        throw std::invalid_argument{"Illegal character in character literal."};
 
    return aa20{}.assign_char(c);
}
 
template <meta::detail::literal_buffer_string str>
constexpr std::vector<aa20> operator""_aa20()
{
    return detail::string_literal<str, aa20>();
}
 
} // namespace literals
 
} // namespace bio::alphabet