aa10li.hpp

Go to the documentation of this file.

// SPDX-FileCopyrightText: 2006-2025 Knut Reinert & Freie Universität Berlin
// SPDX-FileCopyrightText: 2016-2025 Knut Reinert & MPI für molekulare Genetik
// SPDX-License-Identifier: BSD-3-Clause

 
#pragma once
 
#include <vector>
 
#include <seqan3/alphabet/aminoacid/aminoacid_base.hpp>
#include <seqan3/alphabet/aminoacid/concept.hpp>
#include <seqan3/utility/char_operations/transform.hpp>
 
namespace seqan3
{
class aa10li : public aminoacid_base<aa10li, 10>
{
private:
    using base_t = aminoacid_base<aa10li, 10>;

    friend base_t;
    friend base_t::base_t;
 
public:
    constexpr aa10li() noexcept = default;                           
    constexpr aa10li(aa10li const &) noexcept = default;             
    constexpr aa10li(aa10li &&) noexcept = default;                  
    constexpr aa10li & operator=(aa10li const &) noexcept = default; 
    constexpr aa10li & operator=(aa10li &&) noexcept = default;      
    ~aa10li() noexcept = default;                                    

    using base_t::base_t;
 
private:
    static constexpr char_type rank_to_char_table[alphabet_size]{'A', 'B', 'C', 'F', 'G', 'H', 'I', 'J', 'K', 'P'};

    static constexpr rank_type char_to_rank(char_type const chr)
    {
        using index_t = std::make_unsigned_t<char_type>;
        return char_to_rank_table[static_cast<index_t>(chr)];
    }

    static constexpr char_type rank_to_char(rank_type const rank)
    {
        return rank_to_char_table[rank];
    }

    static constexpr std::array<rank_type, 256> char_to_rank_table{
        []() constexpr
        {
            std::array<rank_type, 256> ret{};
 
            // initialize with 'A' because S appears most frequently and gets converted to A in this alphabet
            ret.fill(0u); // Value-initialisation of std::array does usually initialise. `fill` is explicit.
 
            // reverse mapping for characters and their lowercase
            for (rank_type rnk = 0u; rnk < alphabet_size; ++rnk)
            {
                ret[static_cast<rank_type>(rank_to_char_table[rnk])] = rnk;
                ret[static_cast<rank_type>(to_lower(rank_to_char_table[rnk]))] = rnk;
            }
 
            ret['D'] = ret['B'];
            ret['d'] = ret['B']; // Convert D to B (either D/N).
            ret['E'] = ret['B'];
            ret['e'] = ret['B']; // Convert E to B (either D/N).
            ret['L'] = ret['J'];
            ret['l'] = ret['J']; // Convert L to J (either I/L).
            ret['M'] = ret['J'];
            ret['m'] = ret['J']; // Convert M to J (either I/L).
            ret['N'] = ret['H'];
            ret['n'] = ret['H']; // Convert N to H.
            ret['O'] = ret['K'];
            ret['o'] = ret['K']; // Convert Pyrrolysine to K.
            ret['Q'] = ret['B'];
            ret['q'] = ret['B']; // Convert Q to B (either D/N).
            ret['R'] = ret['K'];
            ret['r'] = ret['K']; // Convert R to K.
            ret['S'] = ret['A'];
            ret['s'] = ret['A']; // Convert S to A.
            ret['T'] = ret['A'];
            ret['t'] = ret['A']; // Convert T to A.
            ret['U'] = ret['C'];
            ret['u'] = ret['C']; // Convert Selenocysteine to C.
            ret['V'] = ret['I'];
            ret['v'] = ret['I']; // Convert V to I.
            ret['W'] = ret['F'];
            ret['w'] = ret['F']; // Convert W to F.
            ret['X'] = ret['A'];
            ret['x'] = ret['A']; // Convert unknown amino acids to Alanine.
            ret['Y'] = ret['F'];
            ret['y'] = ret['F']; // Convert Y to F.
            ret['Z'] = ret['B'];
            ret['z'] = ret['B']; // Convert Z (either E/Q) to B (either D/N).
            ret['*'] = ret['F']; // The most common stop codon is UGA. This is most similar to a Tryptophan which in
                                 // this alphabet gets converted to Phenylalanine.
 
            return ret;
        }()};
};
 
// ------------------------------------------------------------------
// containers
// ------------------------------------------------------------------

using aa10li_vector = std::vector<aa10li>;
 
// ------------------------------------------------------------------
// literals
// ------------------------------------------------------------------
inline namespace literals
{

constexpr aa10li operator""_aa10li(char const c) noexcept
{
    return aa10li{}.assign_char(c);
}

SEQAN3_WORKAROUND_LITERAL aa10li_vector operator""_aa10li(char const * const s, size_t const n)
{
    aa10li_vector r;
    r.resize(n);
 
    for (size_t i = 0; i < n; ++i)
        r[i].assign_char(s[i]);
 
    return r;
}

 
} // namespace literals
 
} // namespace seqan3