- DNA, RNA, and protein sequence representations are fundamental to bioinformatics because they provide standardized ways to describe biological molecules in digital form. These representations allow computers to store, retrieve, compare, analyse, and interpret biological information. DNA and RNA sequences are represented as ordered arrangements of nucleotide symbols, whereas protein sequences are represented as ordered arrangements of amino acid symbols. Although all three are biological polymers, they differ in their molecular composition, biological functions, alphabets, structural characteristics, and computational applications. Understanding their representation is essential for genome analysis, transcriptomics, proteomics, sequence alignment, gene prediction, molecular evolution, and functional annotation.
- DNA (deoxyribonucleic acid) is the primary hereditary material in most organisms. It stores genetic information in the form of nucleotide sequences that contain instructions for the development, functioning, and reproduction of living systems. A DNA nucleotide consists of a deoxyribose sugar, a phosphate group, and one of four nitrogenous bases: adenine (A), thymine (T), cytosine (C), or guanine (G). In bioinformatics, DNA sequences are conventionally represented using the four letters A, T, C, and G. A sequence such as ATGCGTACGTTAG represents the order of nucleotide bases along a DNA strand. The digital representation does not directly describe the complete three-dimensional structure of DNA, but it provides the primary sequence information needed for computational analysis.
- DNA is generally double-stranded, with two complementary strands held together by hydrogen bonding between their bases. Adenine pairs with thymine, while cytosine pairs with guanine. The two strands have opposite orientations, known as antiparallel orientation: one runs from 5′ to 3′, and the other from 3′ to 5′. Sequence files usually represent a single strand in the 5′ to 3′ direction. When analysing genomic regions, researchers must therefore consider both the forward strand and its reverse complement. Genes can be located on either DNA strand, and their orientation determines how their sequences are interpreted during transcription and gene annotation.
- The complementary DNA sequence is obtained by replacing each nucleotide with its paired base. The reverse complement is obtained by reversing the complementary sequence so that it is also written in the conventional 5′ to 3′ direction. For example, the reverse complement of ATGCCG is CGGCAT. Reverse-complement operations are important in genome searching, primer design, gene prediction, sequence alignment, and read mapping. Many bioinformatics tools automatically examine both orientations when searching for sequence matches.
- DNA sequence representation may include symbols beyond the four standard nucleotides. These symbols are defined by the International Union of Pure and Applied Chemistry (IUPAC) nucleotide nomenclature and represent uncertain or ambiguous positions. N represents any nucleotide, R represents A or G, Y represents C or T, S represents G or C, and W represents A or T. Other symbols represent additional combinations of possible bases. These codes are useful when sequencing results contain uncertain positions, when a reference sequence represents known variation, or when a consensus sequence is generated from multiple samples. They preserve uncertainty rather than assigning an unsupported nucleotide.
- DNA sequences are used in many forms of computational analysis. Genomic DNA sequences may represent complete chromosomes, scaffolds, contigs, genes, exons, introns, promoters, enhancers, and other regulatory regions. Mitochondrial and chloroplast DNA sequences are also represented using the same nucleotide alphabet, although their genome organization and biological characteristics differ from nuclear genomes. DNA sequence analysis supports gene identification, genome assembly, variant detection, comparative genomics, population genetics, evolutionary studies, and the identification of genetic markers.
- RNA (ribonucleic acid) is another nucleotide-based biological molecule. It plays important roles in gene expression, protein synthesis, regulation, catalysis, and cellular processes. RNA is generally single-stranded, although it can fold into complex secondary and three-dimensional structures. An RNA nucleotide contains ribose sugar, a phosphate group, and one of four nitrogenous bases: adenine (A), uracil (U), cytosine (C), or guanine (G). RNA sequence representations therefore use the alphabet A, U, C, and G. Uracil replaces thymine, which is found in DNA. An example of an RNA sequence is AUGCGUACGUUAG.
- RNA is commonly produced by transcription from a DNA template. During transcription, RNA polymerase synthesizes an RNA strand complementary to the DNA template strand, using uracil instead of thymine. The resulting RNA sequence corresponds to the coding DNA strand, except that U replaces T, assuming the sequences are written in the same orientation and represent the same transcribed region. RNA sequences may undergo processing and modification before becoming mature functional molecules. In eukaryotes, messenger RNA commonly undergoes 5′ capping, intron removal through splicing, and 3′ polyadenylation. These processes influence the final transcript sequence and its representation in bioinformatics databases.
- Different types of RNA have distinct biological roles and are represented using the same nucleotide alphabet. Messenger RNA (mRNA) carries protein-coding information from DNA to ribosomes. Transfer RNA (tRNA) transports amino acids during translation and contains structural features that support codon recognition. Ribosomal RNA (rRNA) forms essential components of ribosomes and contributes to protein synthesis. MicroRNA (miRNA) and small interfering RNA (siRNA) participate in gene regulation, while long non-coding RNA (lncRNA) molecules have diverse regulatory and structural functions. Other RNA categories include small nuclear RNA, small nucleolar RNA, circular RNA, and various classes of non-coding transcripts.
- RNA sequence representation may also contain ambiguity symbols, particularly in transcriptome and sequencing datasets. The IUPAC nucleotide codes used for DNA are also applicable to RNA, with U replacing T where appropriate. The symbol N represents an uncertain nucleotide, while other ambiguity symbols represent possible combinations of bases. RNA sequence records may additionally contain modified nucleotides, but standard sequence alphabets do not represent every chemical modification directly. Modifications such as methylation, pseudouridine formation, and other RNA modifications may require separate annotation or specialized data formats.
- RNA molecules can form secondary structures through intramolecular base pairing. These structures include stems, hairpin loops, internal loops, bulges, and multi-branch junctions. RNA sequence analysis may therefore combine the primary nucleotide sequence with predicted or experimentally determined structural information. RNA secondary structure can be represented using dot-bracket notation, in which dots represent unpaired nucleotides and matching parentheses represent paired positions. For example, a simplified dot-bracket representation might look like (((…))). More complex structures may require additional bracket types or specialized notation. RNA structure information is important in the analysis of ribosomal RNA, transfer RNA, regulatory RNA, and other functional transcripts.
- Protein sequences represent the linear arrangement of amino acids in a polypeptide chain. Proteins are synthesized through translation of messenger RNA, in which ribosomes interpret nucleotide triplets called codons and incorporate the corresponding amino acids. A protein sequence is conventionally represented using a standard alphabet of 20 amino acids. Each amino acid has a three-letter abbreviation and a one-letter code. For example, alanine is represented by Ala or A, glycine by Gly or G, and lysine by Lys or K. A protein sequence such as MKWVTFISLLFLFSSAYS represents the ordered amino acid residues of a polypeptide.
- The standard one-letter amino acid codes are A for alanine, R for arginine, N for asparagine, D for aspartic acid, C for cysteine, E for glutamic acid, Q for glutamine, G for glycine, H for histidine, I for isoleucine, L for leucine, K for lysine, M for methionine, F for phenylalanine, P for proline, S for serine, T for threonine, W for tryptophan, Y for tyrosine, and V for valine. These codes provide a compact and standardized representation that is widely used in protein databases, sequence alignment tools, structural analysis programs, and computational biology applications.
- Protein sequence alphabets also include special symbols. B may represent aspartic acid or asparagine, while Z may represent glutamic acid or glutamine. X is commonly used when the identity of an amino acid is unknown or cannot be determined. U represents selenocysteine, and O represents pyrrolysine. The asterisk is often used to indicate a translation termination signal in a translated sequence, although a stop codon is not an amino acid and is not part of the mature protein chain. A hyphen is commonly used to represent a gap in an aligned protein sequence. These symbols are important when handling uncertain residues, non-standard amino acids, translated sequences, and multiple sequence alignments.
- Protein sequence representation is closely connected to protein structure. The primary structure of a protein is its amino acid sequence, while secondary structure describes local arrangements such as alpha helices and beta sheets. Tertiary structure describes the overall three-dimensional folding of an individual polypeptide chain, and quaternary structure describes the organization of multiple polypeptide subunits. A linear protein sequence provides the starting information for structural analysis, but it does not directly display the complete three-dimensional conformation. Structural prediction and experimental methods are used to investigate how amino acid sequences determine protein folding, stability, interactions, and biological activity.
- Protein sequences may also contain information about post-translational modifications, such as phosphorylation, glycosylation, acetylation, methylation, and ubiquitination. These modifications can influence protein activity, localization, stability, and interactions. They are not always represented directly in a basic one-letter amino acid sequence. Instead, modification information is often stored as annotations associated with specific amino acid positions or represented using specialized proteomics data formats. This distinction is important because two protein molecules with the same amino acid sequence may differ in their chemical modifications and biological properties.
- The relationship between DNA, RNA, and protein sequences is described by the central dogma of molecular biology. Genetic information is generally transferred from DNA to RNA through transcription and from RNA to protein through translation. A DNA coding sequence is transcribed into a corresponding RNA sequence, with uracil replacing thymine. The RNA sequence is then interpreted in groups of three nucleotides, called codons, to determine the amino acid sequence. The genetic code defines the relationship between codons and amino acids. The standard genetic code contains 64 codons, including codons specifying amino acids and three principal termination codons. Alternative genetic codes occur in some organisms and organelles, so the appropriate genetic code must be selected during computational translation.
- An illustrative example of DNA transcription and translation is shown below:
Coding DNA: 5′-ATG GCT TTT GAA TAG-3′
mRNA: 5′-AUG GCU UUU GAA UAG-3′
Protein: Met Ala Phe Glu Stop
- The stop codon signals the termination of translation and is not incorporated as an amino acid into the resulting protein. The example assumes that the DNA sequence is a complete coding sequence written in the correct reading frame. Real biological sequences may contain introns, untranslated regions, alternative transcript forms, incomplete coding regions, and other features that must be considered before translation.
- Reading frames are important when converting nucleotide sequences into protein sequences. Because codons consist of three nucleotides, a nucleotide sequence can be read in three possible frames on one strand. Double-stranded DNA provides three additional reading frames on the opposite strand, giving six possible reading frames. Gene prediction and open reading frame analysis use these possibilities to identify candidate protein-coding regions. An open reading frame is a stretch of nucleotide sequence that can be translated without encountering an in-frame stop codon, although the presence of an open reading frame alone does not prove that a region is a functional gene.
- Biological sequence representations are stored in standardized file formats to support data exchange and computational analysis. FASTA is one of the most widely used formats for DNA, RNA, and protein sequences. Each FASTA record begins with a header line starting with the greater-than symbol, followed by an identifier and optional descriptive information. The sequence is written on subsequent lines. FASTA is simple and widely supported, but it does not normally contain per-position sequencing quality scores or a standardized feature annotation structure.
- An example of a FASTA file containing DNA, RNA, and protein records is shown below:
>DNA_sequence_1 Example DNA sequence
ATGCGTACGTTAGCTAG
>RNA_sequence_1 Example RNA sequence
AUGCGUACGUUAGCUAG
>Protein_sequence_1 Example protein sequence
MKWVTFISLLFLFSSAYS
- FASTQ is particularly important for storing sequencing reads because it includes both the nucleotide sequence and a quality score for each base. A FASTQ record generally contains four lines: a sequence identifier beginning with @, the nucleotide sequence, a separator line beginning with +, and a string of encoded quality values. These quality values are commonly represented using Phred scores, which express the estimated probability that a base call is incorrect. FASTQ is primarily used for nucleotide sequencing data rather than ordinary protein sequence records.
- An illustrative FASTQ record is shown below:
@read_001
ATGCGTACGTTAG
+
IIIIIIIIIIIII
- The quality string contains encoded scores, and its interpretation depends on the applicable encoding convention. Researchers should verify the encoding used by their sequencing dataset before performing quality assessment or filtering.
- Other formats provide more detailed representations of biological sequences and their annotations. GenBank and EMBL flat-file formats store nucleotide sequences together with features such as genes, coding sequences, exons, introns, regulatory regions, and bibliographic information. GFF3 and GTF files describe genomic features and their coordinates, including genes, transcripts, exons, and coding regions. These annotation files are generally linked to a reference genome or sequence record rather than replacing the underlying nucleotide sequence. Protein records may be stored in FASTA or richer database formats containing functional annotations, accession identifiers, domains, and evidence information.
- Specialized structural formats are used when sequence information must be connected to three-dimensional molecular structures. Protein Data Bank (PDB) and PDBx/mmCIF formats describe atomic coordinates and structural metadata for proteins, nucleic acids, and molecular complexes. These formats are not simply alternative ways of writing a linear sequence; they provide structural information that can be linked to the corresponding biological sequence. Structural records may include residue numbering, chain identifiers, modified residues, ligands, and experimental information. Sequence-to-structure mapping is therefore important when relating protein sequence positions to experimentally determined or predicted structural features.
- Biological sequence representation also requires the use of stable identifiers and metadata. Sequence identifiers distinguish records and allow researchers to retrieve, compare, and reference biological sequences. Accession numbers assigned by databases such as GenBank, ENA, DDBJ, RefSeq, and UniProt provide persistent references to sequence records, although records may be revised or updated. Metadata may include organism name, gene name, sequence length, molecule type, biological source, database version, and evidence supporting the annotation. Such information is essential for distinguishing similar sequences from different organisms, identifying reference records, and reproducing analyses.
- Computational libraries provide convenient methods for representing and manipulating biological sequences. Biopython supports DNA, RNA, and protein sequence objects through its Seq and SeqRecord classes. These objects support sequence operations and can store identifiers, descriptions, and annotations. The following example demonstrates how DNA, RNA, and protein sequences can be represented and examined using Biopython.
from Bio.Seq import Seq
from Bio.SeqRecord import SeqRecord
dna = Seq(“ATGGCTTTTGAATAG”)
rna = dna.transcribe()
protein = dna.translate()
dna_record = SeqRecord(
dna,
id=”DNA_001″,
description=”Example coding DNA sequence”
)
rna_record = SeqRecord(
rna,
id=”RNA_001″,
description=”Example messenger RNA sequence”
)
protein_record = SeqRecord(
protein,
id=”PROTEIN_001″,
description=”Translated protein sequence”
)
print(dna_record)
print(rna_record)
print(protein_record)
- The same library can read and write sequence records in common formats. For example, the following code reads a FASTA file and reports the identifier and sequence length for each record.
from Bio import SeqIO
for record in SeqIO.parse(“sequences.fasta”, “fasta”):
print(“Identifier:”, record.id)
print(“Description:”, record.description)
print(“Length:”, len(record.seq))
print(“Sequence:”, record.seq)
- Sequence representations can also be converted between formats when required. For example, FASTA records can be written from parsed sequence objects, while annotated GenBank records can be converted to other supported formats. Such conversions should be performed carefully because some formats preserve more metadata than others. Converting an annotated GenBank record into basic FASTA, for instance, may retain the sequence and header but discard detailed feature annotations.
- In large-scale bioinformatics, sequences are not always stored as ordinary character strings. Efficient data structures are required for genome-scale analysis, sequence searching, alignment, and read mapping. Arrays, hash tables, suffix arrays, FM-indexes, and k-mer tables support different computational operations. Compressed representations and indexed file formats reduce storage requirements and allow rapid access to specific regions of large sequence datasets. For example, indexed FASTA files can support retrieval of selected genomic regions, while BAM and CRAM formats provide compact storage and indexed access to sequence-read alignments.
- The representation of DNA, RNA, and protein sequences must also account for biological uncertainty and data quality. Sequencing errors may introduce incorrect nucleotides, ambiguous positions, or missing information. Transcript sequences may contain alternative splice forms, while protein sequences may be incomplete or contain uncertain residues. Reference sequences may represent a particular individual, strain, or assembly rather than the complete variation found within a species. Researchers should therefore distinguish between experimentally observed sequences, consensus sequences, predicted sequences, and database-derived reference sequences.
- Quality control and validation help ensure that sequence representations are suitable for downstream analysis. DNA and RNA records should be checked for valid nucleotide symbols, correct orientation, sequence length, and possible contamination. Protein records should be examined for unexpected characters, ambiguous residues, and inappropriate termination symbols. FASTQ records require additional checks of sequence and quality-string lengths and quality-score encoding. Tools such as FastQC, fastp, Biopython, and other format-validation utilities support these procedures. Proper validation reduces errors that could otherwise affect alignment, gene prediction, variant detection, and functional interpretation.
- DNA, RNA, and protein sequence representations are closely integrated in modern bioinformatics workflows. Genome analysis begins with DNA sequence representation and often proceeds through gene prediction, transcript identification, and functional annotation. Transcriptomics examines RNA sequences and their expression, while proteomics investigates proteins and their modifications, abundance, and interactions. Comparative genomics and molecular evolution use sequence representations to identify conserved regions and evolutionary changes. Structural bioinformatics connects protein and nucleic acid sequences with their molecular structures, while clinical and agricultural genomics use sequence data to investigate disease, inherited traits, and biological variation.
- Reproducibility requires researchers to record the sequence source, accession number, organism, reference version, file format, software version, and processing parameters. The original sequence files should be preserved whenever possible, and any changes made during trimming, filtering, conversion, or annotation should be documented. Database records may change over time, so recording accession versions and retrieval dates can be important. These practices help ensure that sequence data can be independently checked and that computational results can be reproduced.
- DNA, RNA, and protein sequence representations provide the basic language through which biological information is stored and analysed computationally. DNA sequences represent hereditary information, RNA sequences represent transcripts and functional RNA molecules, and protein sequences describe the amino acid chains responsible for many cellular activities. Their standardized alphabets, sequence formats, identifiers, annotations, and computational data structures make it possible to integrate molecular information across different levels of biological organization. A clear understanding of these representations is essential for accurate sequence analysis and for nearly every major application of modern bioinformatics.