diff --git a/TAOK3.py b/TAOK3.py
new file mode 100644
index 0000000..2866805
--- /dev/null
+++ b/TAOK3.py
@@ -0,0 +1,376 @@
+import os
+from abc import ABC, abstractmethod
+from typing import Optional
+
+import numpy as np
+from Bio import SeqIO, SeqRecord, SeqUtils
+
+
+def is_in_gc_bounds(seq_record: SeqRecord, gc_bounds: tuple) -> bool:
+    """
+    Check if the sequence is in the range of GC-content bounds
+
+    Arguments:
+    - seq_record(SeqRecord): the sequence object to check
+    - gc_bounds(tuple): contain min and max GC-content bounds
+
+    Return:
+    - bool: the result of the check
+    """
+    gc_min, gc_max = gc_bounds[0], gc_bounds[1]
+    gc_content = SeqUtils.GC123(seq_record.seq)[0]
+    return gc_min <= gc_content <= gc_max
+
+
+def is_in_length_bounds(seq_record: SeqRecord, length_bounds: tuple) -> bool:
+    """
+    Check if the sequence is in the range of length bounds
+
+    Arguments:
+    - seq(SeqRecord): the sequence object to check
+    - length_bounds(tuple): contain min and max length bounds
+
+    Return:
+    - bool: the result of the check
+    """
+    length_min, length_max = length_bounds[0], length_bounds[1]
+    return length_min <= len(seq_record) <= length_max
+
+
+def is_above_quality_threshold(seq_record: SeqRecord,
+                               quality_threshold: float) -> bool:
+    """
+    Check if the mean of the sequence quality values in FASTQ exceeds
+    the quality threshold of the interest
+
+    To convert quality values, the function uses phred+33
+
+    Arguments:
+    - seq_record(Bio.SeqRecord): the sequence object to check
+    - quality_threshold(int or float): the quality threshold of the interest
+
+    Return:
+    - bool: the result of the check
+    """
+    quality = np.mean(seq_record.letter_annotations["phred_quality"])
+    return quality >= quality_threshold
+
+
+def filter_fastq(input_path: str,
+                 output_filename: Optional[str] = None,
+                 gc_bounds=(0, 100),
+                 length_bounds=(0, 2**32),
+                 quality_threshold=0) -> None:
+    """
+    Main function to select reads in FASTQ format according
+    to three main requirements:
+
+    -correspondence to the range of GC-content bounds
+    The range of GC-content bounds is determined with gc_bounds argument
+
+    -correspondence to the range of length bounds
+    The range of length bounds is determined with length_bounds argument
+
+    -exceeds the quality threshold of the interest
+    The quality threshold is determined with quality_threshold argument
+
+    Function takes the path to the file in input_path argument.
+    Use files with fastq extension only.
+
+    Function output the result of checking in a file that is named according
+    to output_filename(Optional).
+
+    The output file also has fastq extension.
+
+    The output file is saved in the 'fastq_filtrator_results' directory.
+
+    If 'fastq_filtrator_results' directory doesn't exist the program creates it
+    in a current directory.
+
+    Without full names use arguments in a certain order
+    Example: filter_fastq(input_path, output_filename, (0,100), (0,200), 0)
+           # filter_fastq(input_path, output_filename, gc_bounds=(0,100),
+             length_bounds=(0,200), quality_threshold=0)
+
+    In case of changing only one argument, provide its full name!
+    Example: filter_fastq(input_path, output_filename,
+                          length_bounds=(50, 100))
+
+    Arguments:
+
+    - input_path(str): the path to the file
+
+    - output_filename(str): the name for output file with obtained result
+    By default output_filename=None
+    Without output_filename argument the output file is named as input file
+    Name without fastq extention is acceptible.
+    Example: output_filename='result'  # 'result.fastq'
+             output_filename='result.fastq'
+
+    - gc_bounds(tuple or int or float): contain min and max GC-content bounds
+    By default gc_bounds=(0,100)
+    If input contains one number the function accepts it as a maximum bound
+    Examples: gc_bounds=(20,40)
+              gc_bounds=40  # (0,40)
+
+    - length_bounds(tuple or int or float): contain mini and max length bounds
+    By default length_bounds=(0,4294967296)
+    If input contains one number the function accepts it as a maximum bound
+    Examples: length_bounds=(10,90)
+              length_bounds=90  # (0,90)
+
+    - quality_threshold(int or float): the quality threshold of the interest
+    By default quality_threshold=0
+    Examples: quality_threshold=10
+
+    There are three functions that are used in the main function:
+
+    - is_in_gc_bounds(seq_record, gc_bounds):
+    Check if the sequence falls in the range of GC-content bounds
+
+    - is_in_length_bounds(seq_record, length_bounds):
+    Check if the sequence falls in the range of length bounds
+
+    - is_above_quality_threshold(seq_record, quality_threshold):
+    Check if the mean of quality values exceeds the quality threshold
+
+    Return:
+    - file: file with fastq extension containing selected fragments.
+
+    For more information please see README
+
+    """
+    if type(length_bounds) is int:
+        length_bounds = 0, length_bounds
+    if type(gc_bounds) is int or type(gc_bounds) is float:
+        gc_bounds = 0, gc_bounds
+    sequences = SeqIO.parse(input_path, "fastq")
+    good_reads = []
+    for seq_record in sequences:
+        if (is_in_gc_bounds(seq_record, gc_bounds) and
+            is_in_length_bounds(seq_record, length_bounds) and
+            is_above_quality_threshold(seq_record, quality_threshold)):
+            good_reads += [seq_record]
+    if len(good_reads) == 0:
+        raise ValueError('There are no sequences suited to requirements')
+    if output_filename is None:
+        input_filename = os.path.split(input_path)[-1]
+        output_filename = input_filename
+    if not (output_filename.endswith('.fastq')):
+        output_filename = output_filename + '.fastq'
+    current_directory = os.getcwd()
+    path = os.path.join(current_directory, 'fastq_filtrator_results')
+    if not (os.path.exists(path)):
+        os.mkdir(path)
+    output_path = os.path.join(path, output_filename)
+    if os.path.exists(output_path):
+        error = 'File with such name exists! Change output_filename arg!'
+        raise ValueError(error)
+    SeqIO.write(good_reads, output_path, "fastq")
+
+
+class BiologicalSequence(ABC):
+    """
+    The abstract class for biological sequences
+    """
+    @abstractmethod
+    def __len__(self):
+        pass
+
+    @abstractmethod
+    def __getitem__(self, item):
+        pass
+
+    @abstractmethod
+    def __str__(self):
+        pass
+
+    def __repr__(self):
+        pass
+
+    @abstractmethod
+    def is_alphabet_correct(self):
+        pass
+
+
+class NucleicAcidSequence(BiologicalSequence):
+    """
+    Class for nucleic acids
+    """
+    def __init__(self, seq):
+        self.seq = seq
+
+    def __len__(self):
+        return len(self.seq)
+
+    def __getitem__(self, item):
+        return self.seq[item]
+
+    def __str__(self):
+        return self.seq
+
+    def __repr__(self):
+        return self.seq
+
+    def is_alphabet_correct(self) -> bool:
+        """
+        The function check does the sequence contain
+        standard A, G, C, T, U nucleotides
+
+        Returns: bool - the result of check
+        """
+        return type(self).is_correct(self)
+
+    def complement(self) -> BiologicalSequence:
+        """
+        Output the complementary sequence
+        The complementarity rule could be found here:
+        https://en.wikipedia.org/wiki/Complementarity_(molecular_biology)
+
+        Arguments:
+        - self: the sequence obj to change
+
+        Return: DNAsequence or RNAsequence - the result sequence object
+        """
+        complement = self.seq.translate(type(self).rule_complement)
+        complement_seq = type(self)(complement)
+        return complement_seq
+
+    def gc_content(self) -> float:
+        """
+        Check the gc content in the sequence object
+        Returns: int - the percentage of GC content
+        """
+        gc_counter = 0
+        for nucl in self.seq:
+            if nucl in ('G', 'C', 'g', 'c'):
+                gc_counter += 1
+        gc_share = gc_counter / len(self.seq) * 100
+        return gc_share
+
+
+class RNASequence(NucleicAcidSequence):
+    """
+    The class for RNA sequences
+    """
+    rule_complement = 'AUCG'.maketrans('AaUuCcGg', 'UuAaGgCc')
+
+    def __init__(self, seq):
+        super().__init__(seq)
+        if not (super().is_alphabet_correct()):
+            raise ValueError('The sequence does not correspond to RNA')
+
+    def is_correct(self) -> bool:
+        """
+        Check if the sequence is RNA
+
+        Arguments:
+        - self: the sequence object to check
+
+        Return:
+        - bool: the result of the check
+        """
+        return set(self.seq) <= set('AaGgCcUu')
+
+
+class DNASequence(NucleicAcidSequence):
+    """
+    The class for DNA sequences
+    """
+    rule_complement = 'ATCG'.maketrans('AaTtCcGg', 'TtAaGgCc')
+    rule_transcription = 'AUCG'.maketrans('Tt', 'Uu')
+
+    def __init__(self, seq):
+        super().__init__(seq)
+        if not (super().is_alphabet_correct()):
+            raise ValueError('The sequence does not correspond to DNA')
+
+    def is_correct(self) -> bool:
+        """
+        Check if the sequence is DNA
+
+        Arguments:
+        - self: the sequence object to check
+
+        Return:
+        - bool: the result of the check
+        """
+        return set(self.seq) <= set('AaGgCcTt')
+
+    def transcribe(self) -> RNASequence:
+        """
+        Transcribe DNA sequence to RNA
+
+        Arguments:
+        - self: the sequence object to change
+
+        Return: str: RNASequence object
+        """
+        transcribe_seq = self.seq.translate(self.rule_transcription)
+        rna_seq = RNASequence(transcribe_seq)
+        return rna_seq
+
+
+class AminoAcidSequence(BiologicalSequence):
+    """
+    The class for amino acid sequences
+    """
+    aa_alphabet = 'ACDEFGHIKLMNPQRSTVWY'
+
+    def __init__(self, seq):
+        self.seq = seq
+        if not (self.is_alphabet_correct()):
+            error = 'The sequence does not match standard protein code'
+            raise ValueError(error)
+
+    def __len__(self):
+        return len(self.seq)
+
+    def __getitem__(self, item):
+        return self.seq[item]
+
+    def __str__(self):
+        return self.seq
+
+    def __repr__(self):
+        return self.seq
+
+    def is_alphabet_correct(self) -> bool:
+        """
+        The function checks if the sequence object contains
+        standard amino acid code
+        Returns: list of alternative frames (AminoAcidSequence)
+        """
+        return set(self.seq.upper()) <= set(self.aa_alphabet)
+
+    def search_for_alt_frames(self, alt_start_aa='M') -> list:
+        """
+        Search for alternative frames in a protein sequences
+        Use only one-letter code
+
+        Search is not sensitive for letter case
+        Without an alt_start_aa argument search for
+        frames that start with methionine ('M')
+        To search frames with alternative start codon
+        add alt_start_aa argument
+
+        The function ignores the last three amino acids in sequences
+
+        Arguments:
+        - self: sequence object to check
+        - alt_start_aa (str): the name of an amino acid that
+        is encoded by alternative start AA (Optional)
+        Default: alt_start_aa='I'
+
+        Return: the list of alternative frames
+        """
+        alternative_frames = []
+        num_position = 0
+        for amino_acid in self.seq[1:-3]:
+            alt_frame = ''
+            num_position += 1
+            if amino_acid.upper() == alt_start_aa:
+                alt_frame += self.seq[num_position:]
+                alt_frame = AminoAcidSequence(alt_frame)
+                alternative_frames.append(alt_frame)
+        return alternative_frames