- Genome assemblies and assembly accessions are fundamental concepts in bioinformatics because they provide organized representations of genomic sequences and unique identifiers for retrieving, comparing, and documenting them. A genome assembly is a computational reconstruction of an organism’s genome from sequencing data, while an assembly accession is a database identifier that refers to a particular assembly and its version. These concepts are essential in genome sequencing, genome annotation, comparative genomics, variant calling, transcriptomics, population genetics, and reproducible biological research.
- Genome sequencing technologies generate numerous DNA sequence fragments, commonly called reads. Because these reads represent only portions of a genome, computational methods are used to reconstruct longer sequences by identifying overlaps, shared sequence patterns, or relationships among reads. This process is known as genome assembly. Depending on the organism, sequencing technology, data quality, and assembly method, the result may contain a collection of short contigs, longer scaffolds, chromosome-level sequences, or a nearly complete genome. Reference-guided assembly uses an existing genome as a framework, whereas de novo assembly reconstructs the genome without requiring a closely related reference sequence.
- A contig is a continuous assembled sequence without internal gaps represented as unknown bases. A scaffold is an ordered and oriented collection of contigs whose relative positions are supported by additional evidence, although gaps may remain between them. Chromosome-level assemblies organize sequences into chromosome representations, while complete genome assemblies aim to represent all chromosomes or genomic molecules with no unresolved gaps under the database’s stated criteria. Assembly quality can also be assessed through sequence accuracy, completeness, contamination, chromosome assignment, and the representation of repetitive or structurally complex regions. A high assembly level does not by itself guarantee that every base is correct or that all biological variation has been captured.
- Genome assemblies are organized into records that describe the assembled sequences and their relationships. An assembly record can include the organism name, strain or individual, assembly name, submitting organization, submission and release dates, sequencing methods, assembly statistics, component sequence identifiers, and associated annotation information. For complex eukaryotic genomes, the assembly may include a primary assembly, organelle sequences, alternate loci, and additional sequence representations. The primary assembly generally provides a non-redundant representation of the main genomic content, while alternate loci can represent alternative forms of particular genomic regions. The exact organization depends on the assembly and the database’s assembly model.
- An assembly accession is a unique identifier assigned to a genome assembly record. At NCBI, the two commonly encountered assembly accession prefixes are GCA_ and GCF_. The GCA_ prefix identifies an assembly represented in the GenBank collection, while GCF_ identifies an assembly represented in the RefSeq collection. Both are followed by nine digits and a version suffix. For example, GCA_000001405.15 identifies a GenBank assembly record for the human reference assembly GRCh38, while GCF_000001405.40 identifies a later RefSeq assembly record for GRCh38. These are examples of assembly identifiers; users should check the current database record to confirm its status and associated files. <Cite refs={[“turn565265search0″,”turn565265search3”]}/>
- GenBank and RefSeq assembly records have different roles. GenBank is a primary archival resource that preserves genome assemblies submitted by researchers and sequencing organizations through the international nucleotide sequence database system. RefSeq provides a selected and standardized collection of reference sequences and may maintain corresponding assembly records derived from GenBank assemblies. A GenBank assembly and its RefSeq counterpart can represent the same underlying assembly while carrying different accession prefixes and independently managed version histories. Therefore, the numerical portions and version suffixes should not be assumed to match in every case. Researchers should verify the relationship between the two records rather than treating GCA_ and GCF_ identifiers as interchangeable. <Cite refs={[“turn565265search1″,”turn565265search2”]}/>
- The version suffix is a particularly important part of an assembly accession. In an identifier such as GCF_000001405.40, the number after the period identifies the version of that assembly record. When sequence content or assembly organization changes in a way that affects the defined assembly, the version is incremented. The accession without its version suffix identifies the assembly record family, but it may not specify the exact sequence content used in an analysis. Metadata-only changes do not necessarily cause an assembly version increase. Researchers should therefore use the full accession and version when reporting an assembly in a publication, workflow, database, or research dataset. <Cite refs={[“turn565265search0″,”turn565265search4”]}/>
- Individual sequences within an assembly also have their own sequence accessions and version numbers. An assembly accession identifies the collection of sequences belonging to an assembly, whereas an individual sequence accession identifies a particular chromosome, contig, scaffold, or other component sequence. For example, a chromosome sequence may have an accession such as CM000663.2, while the assembly containing that chromosome has a separate GCA_ or GCF_ accession. This distinction matters because a change to a component sequence can lead to an assembly version update. A complete record of an analysis may therefore require the assembly accession, individual sequence identifiers when relevant, and the annotation release used to interpret genomic features. <Cite refs={[“turn565265search2″,”turn565265search4”]}/>
- Genome assembly accessions are associated with assembly names and commonly used labels. For example, the human reference assembly is widely known as GRCh38, and it is often called hg38 in the UCSC Genome Browser. The mouse assembly GRCm38 is also known as mm10 in the UCSC environment. Such names are useful for communication, but they do not always uniquely identify a particular assembly version or annotation release. A genome may have multiple patch releases, assembly updates, or database-specific representations. The most reliable way to identify the exact sequence collection is to record the full assembly accession and version together with the assembly name and, where appropriate, the annotation release. <Cite refs={[“turn565265search3″,”turn565265search9”]}/>
- Genome assembly records also have status information. In NCBI’s assembly system, a version may be designated as latest, replaced, or suppressed. A latest version is the current version of that assembly record. A replaced version has been superseded by a newer version, but its historical sequence and associated files may remain accessible. A suppressed assembly has been removed from ordinary search or comparative-analysis results and may be retrievable only through specific identifiers, depending on the reason for suppression. Status information should be checked before choosing an assembly, especially when building a new pipeline or reproducing an older analysis. A replaced assembly is not necessarily scientifically unusable; it may be required when reproducing published work that used that exact version. <Cite refs={[“turn565265search0″,”turn565265search11”]}/>
- Assembly quality is commonly described using several complementary metrics. Total assembly length indicates the amount of assembled sequence, while the number of contigs and scaffolds gives an indication of fragmentation. The N50 statistic is the sequence length at which 50% of the total assembly length is contained in contigs or scaffolds of that length or longer. L50 represents the number of such sequences needed to account for half of the total assembly length. Larger N50 values and smaller L50 values often indicate greater continuity, but these metrics do not directly measure sequence correctness or completeness. Additional measures, such as BUSCO completeness, read-mapping rates, k-mer-based estimates, contamination checks, and comparisons with expected genome size, provide further evidence about assembly quality. No single statistic is sufficient to establish that an assembly is biologically accurate.
- Reference genome assemblies are particularly important because they provide coordinate systems for locating genes, regulatory regions, and genetic variants. During read mapping, sequencing reads are aligned to the chosen reference assembly. Variant-calling software then identifies differences between the reads and reference sequence, including single-nucleotide variants and small insertions or deletions. Genome annotation resources assign genes and transcripts to genomic coordinates, while tools such as Ensembl Variant Effect Predictor and SnpEff use these annotations to predict potential variant consequences. If an analysis uses the wrong assembly, reads may map incorrectly, genomic coordinates may refer to different locations, and variant annotations may become unreliable. Assembly identity must therefore be consistent across the relevant stages of a bioinformatics workflow.
- Genome assembly accessions are also useful for data retrieval and automated analysis. NCBI Datasets allows researchers to search for genome assemblies and download associated sequence files, annotations, and metadata. The command-line interface supports retrieval using an assembly accession, a taxonomic name or identifier, or a BioProject accession. For example, a researcher can download a genome data package using a specific assembly accession:
- datasets download genome accession GCF_000001405.40 –filename human_GRCh38_dataset.zip
- This command requests a genome data package for the specified RefSeq assembly. The contents depend on the selected options and the files available for that assembly. Researchers should inspect the resulting metadata and file manifest to confirm that the downloaded data correspond to the intended assembly and annotation. <Cite refs={[“turn565265search8”]}/>
- Genome assemblies and accessions also support data integration across biological databases. Genome browsers, gene databases, sequence archives, and variant resources use assembly identifiers to establish relationships between genomic sequences and their annotations. This enables researchers to combine gene expression measurements, epigenomic signals, sequence variants, and functional annotations in a common coordinate framework. However, datasets may use different assemblies or chromosome naming conventions. Converting genomic coordinates between assemblies, often called liftover, can help align datasets, but some regions cannot be mapped unambiguously, and structural differences may prevent reliable conversion. Coordinate conversion should therefore be documented and validated rather than assumed to be exact.
- For reproducible research, the full assembly accession and version should be recorded alongside the organism, strain or sample when applicable, annotation release, software versions, and relevant analysis parameters. Researchers should retain downloaded files or use reliable archival sources, record checksums where appropriate, and distinguish sequence changes from annotation changes. If an assembly has been replaced, the historical version may still be necessary to reproduce earlier results. When a newer assembly is adopted, analyses involving coordinates, gene models, or variant effects may need to be repeated and compared systematically.
- Genome assemblies and assembly accessions provide the foundation for organizing genomic information in a traceable and reproducible manner. Assembly methods reconstruct genomes from sequencing data, while assembly records describe the resulting sequences and their metadata. Accession identifiers, version suffixes, and status information allow researchers to retrieve the intended sequence collection and distinguish it from previous or alternative releases. Understanding the difference between an assembly accession, an individual sequence accession, an assembly name, and an annotation version is essential for accurate genome analysis, reliable data exchange, and reproducible bioinformatics research.