# Ensembl Gene Page Sections: Complete Guide The Ensembl gene page sections are the labeled blocks and tabs through which the [Ensembl genome browser](/blog/guides/ensembl-genome-browser) presents everything it knows about one gene. The default view opens on a summary header and a genomic location graphic, and the remaining sections (transcripts, variation, expression, phenotypes, homology, regulation, and external references) open as tabs from that header. This page is the single most visited entry point into Ensembl, and most first-time users lose time because they do not know which section answers which question. Once you can map a question ("which transcript is longest?", "is this variant in a coding exon?", "does this gene have an ortholog in mouse?") to a named section, the rest of the browser becomes predictable. This guide walks through every major section of the Ensembl gene page, explains the columns and flags you will actually read, gives worked examples with real genes, and covers the practical mistakes that cost students hours. ## What Are the Ensembl Gene Page Sections, Explained Ensembl is a joint project of EMBL-EBI and the Wellcome Sanger Institute that maintains genome annotation, meaning the computational assignment of genes, transcripts, exons, proteins, variants, and regulatory features onto assembled reference genome sequences. When you search for a gene at ensembl.org, the browser routes you to a gene page whose URL follows the pattern `https://www.ensembl.org//Gene/Summary?g=`. For the human gene BRCA1, the address is built around the stable identifier ENSG00000012048. The page is organized around a fixed skeleton: 1. A **header block** with the gene name, stable ID, biotype, source database, synonyms, and quick-access icons. 2. An **About this gene** description and a **Genomic location** strip showing chromosome, coordinates, and assembly. 3. A row of **section tabs** (the exact set depends on the species and the data available): Location, Transcripts, Expression, Phenotypes, Gene tree, Homology (orthologues), Variation, Regulation, and Literature or external references. 4. Contextual **summary tables and graphics** inside each tab, such as the transcript table, the variant table, and the orthologue table. Two practical facts explain the design. First, everything on the page is anchored to a **stable identifier**, an ID of the form ENSG followed by eleven digits for human genes (for example ENSG00000012048 for BRCA1, or ENSG00000141510 for TP53). Mouse genes carry the prefix ENSMUSG, cow ENSBTAG, chicken ENSGALG, and so on. Second, the sections are cross-linked. Clicking an exon in the transcript graphic jumps you to the sequence view. Clicking a variant in the variation table opens a variant page that links back to the gene. Learning the gene page therefore teaches you the navigation logic of the whole Ensembl genome browser. ### The Two Page Versions You May Encounter Ensembl redesigned its web interface in 2023. Older screenshots and tutorials show a page with a green gene bar, a left-hand "content" menu with dozens of expandable links, and a location graphic called "Region in detail." The current page uses a cleaner header, horizontally arranged tabs, and a streamlined transcript table. The underlying data are the same, the stable IDs are the same, and every section in the old interface still exists in the new one. If a tutorial tells you to click a link that does not exist, look for the same content under the corresponding tab described here. ## Summary Table: Gene Page Sections at a Glance | Section (tab or block) | What it shows | Typical questions it answers | |, - |, - |, - | | Header and summary | Name, stable ID, biotype, source, synonyms, assembly | "What gene is this and which annotation is it?" | | Genomic location | Chromosome, start and end coordinates, cytogenetic band | "Where does this gene sit on GRCh38?" | | Location view | Contig-level graphic, neighboring genes, sequence, basepair view | "What is adjacent to this gene? What does the sequence look like?" | | Transcripts | Transcript table, exon-intron graphics, cDNA and protein sequences | "Which transcripts exist, which is canonical, how many exons?" | | Variation | Variant table with consequences, links to variant pages and frequencies | "Which known variants fall in this gene and what do they do?" | | Phenotypes | Phenotype and disease associations from OMIM, GWAS Catalog, mouse resources | "Is this gene linked to any trait or disease?" | | Expression | Baseline expression by tissue, single-cell summaries where available | "Where is this gene expressed?" | | Gene tree and Homology | Orthologues, paralogues, phylogenetic tree from Ensembl Compara | "What is the mouse or zebrafish counterpart of this gene?" | | Regulation | Regulatory features, motifs, accessible chromatin where available | "Which regulatory elements overlap this locus?" | | Literature | Curated publication list via Europe PMC | "What are the key papers for this gene?" | | External references | Cross-links to RefSeq, UniProt, HGNC, OMIM and others | "What are the IDs for this gene in other databases?" | ### Terms the Gene Page Is Often Confused With | Term | What it actually is | How it differs | |, - |, - |, - | | Ensembl gene page | A page for one gene, summarizing all transcripts | Transcript pages (ENST IDs) describe one transcript only | | Ensembl canonical | Ensembl's own default transcript choice per gene | Not identical to MANE Select, the joint Ensembl and RefSeq choice for human | | MANE Select | A matched pair of Ensembl and RefSeq transcripts agreed for clinical use | One per human protein-coding gene, flagged in the transcript table | | Ensembl release | The version number of the annotation (for example release 110) | Different from the assembly version (for example GRCh38.p14), which describes the underlying sequence | | Stable ID version suffix | The number after the dot in ENSG00000012048.5 | Increments when the annotation of that object changes, not with every release | ## The Header and Gene Summary Block Open any gene page and the top block answers four questions before you scroll. **What is this gene called and what does it do?** The header shows the primary gene symbol (BRCA1), the full name (breast cancer type 1 susceptibility protein in current HGNC nomenclature), and the biotype. The biotype is Ensembl's classification of the gene, with protein_coding the most common value, and others including lncRNA, miRNA, pseudogene, and snoRNA. The "About this gene" paragraph underneath gives a short description assembled from the linked source databases. **What is the stable ID and what version is it?** The identifier appears as, for example, ENSG00000012048.5. The number after the dot is the version. When the annotation of that transcript set changes, the version increments and older versions remain resolvable through the archive site at archive.ensembl.org. This is central to reproducibility: a methods section that cites "BRCA1, ENSG00000012048.5, release 110" can be reproduced exactly, while "BRCA1" alone cannot. **Which database supplied the name and symbol?** Ensembl does not invent human gene names. The header lists the source, usually HGNC (HUGO Gene Nomenclature Committee) for human, MGI (Mouse Genome Informatics) for mouse, and the relevant nomenclature body for other species. Synonyms from the same source appear here too, which matters because the literature freely mixes old and new symbols. **Where is it and what is nearby?** The header carries two miniature graphics, one showing the gene's position on its chromosome with neighboring genes, and one showing its transcript structure at a glance. These are live links. Clicking either jumps into the corresponding section rather than making you find it in a menu. Practical buttons sit in this header: a **Download GTF** button that fetches the gene's annotation in Gene Transfer Format, and an **Export data** button that opens a dialog for FASTA sequences, coordinate data in GFF3 or BED-style output, and other formats. If you only remember one workflow, remember this one: find the gene, press Download GTF, and you have machine-readable annotation for your pipeline in seconds. ## The Location Section: Reading the Genomic Neighborhood Click the **Location** tab and you move from the gene summary to the genome browser proper. The default view is a contig-level graphic showing the gene in red or orange, neighboring genes above and below, and a coordinate axis. From here you can do four things that come up constantly in practice. **Zoom and navigate.** Buttons and a coordinate box let you zoom out to the whole chromosome or zoom in to individual exons. At high magnification the view switches to the **basepair view**, which renders the actual reference sequence with start and stop codons, splice sites, and (where enabled) six-frame translation. This is the fastest way to answer questions like "what is the sequence at the splice donor of exon 2?" without writing a script. **Compare against other tracks.** The "Configure this page" or track menu lets you overlay tracks such as regulatory features, variants, or markers. For human, tracks from the ENCODE project and from the 1000 Genomes Project are available, and for mouse, annotation from the MOUSE (Mutant Mouse Resource) project lines up alongside the core gene models. **Check the assembly.** The location view names the assembly explicitly. Human annotation is currently on GRCh38 (Genome Reference Consortium human build 38, including patch releases such as GRCh38.p14), and mouse is on GRCm39. Every coordinate you read on this page is meaningless without that assembly label. A statement that a variant sits at "chromosome 17, position 43,044,295" refers to one assembly only, and converting between GRCh37 and GRCh38 coordinates requires a liftover tool, not arithmetic, because the assemblies differ by millions of inserted, deleted, and reordered bases. **Export the view.** The export menu produces publication-quality SVG or PNG images of the region, plus the underlying data. A concrete example: the human beta-globin gene HBB (ENSG00000244734) sits on chromosome 11 within the globin locus. Zooming out in the location view shows the neighboring globin genes arranged in the order they are expressed during development, which is exactly the kind of structural observation this section is designed to make visible at a glance. ## The Transcripts Section: The Heart of the Gene Page For most students, the transcripts section is where the real work happens. Click the **Transcripts** tab and you get two things, a table of transcripts and a graphic showing each transcript aligned to the genome. ### Reading the Transcript Table Each row is one transcript with its own stable ID (ENST followed by digits, for example ENST00000357654 for a well-characterized BRCA1 transcript). The columns you will use most are: - **Biotype.** Most transcripts are protein_coding, but a single gene locus commonly produces protein_coding, nonsense_mediated_decay, retained_intron, and lncRNA transcripts simultaneously. BRCA1, for example, lists over a dozen transcripts spanning several biotypes, while the beta-globin gene HBB carries a short, cleanly annotated set dominated by one protein-coding transcript of three exons. - **Protein length.** Shown in amino acids for protein-coding transcripts. BRCA1's canonical protein is 1,863 amino acids, a number you can read directly off the table rather than reconstructing from exons. - **Flags.** Compact badges summarize annotation status. The most important are **Ensembl canonical** (Ensembl's chosen representative transcript, generally the longest conserved exon structure with APPRIS priority), **MANE Select** (the jointly agreed Ensembl and RefSeq match, one per human protein-coding gene, intended as the default for clinical reporting and variant nomenclature), **CCDS** (presence in the Consensus CDS set shared between Ensembl, RefSeq, and UniProt), **TSL** (Transcript Support Level, a 1-to-5 scale reflecting the strength of underlying evidence, with TSL1 being strongest), and, for some species, **GENCODE basic**. - **RefSeq match.** The corresponding NCBI RefSeq transcript, so you can reconcile your Ensembl-based analysis with RefSeq-based resources. ### The Exon-Intron Graphic Below the table, each transcript is drawn with boxes for exons and thin lines for introns. Coding portions of exons are filled more heavily than untranslated regions. Hovering over an exon reports its coordinates, and clicking an exon opens its entry. This graphic is the quickest visual check of [alternative splicing](/blog/guides/alternative-splicing): you can see at a glance that two transcripts share the first five exons but differ in the terminal exon, or that one transcript skips an internal exon entirely. ### Sequences on the Transcript Page Clicking any transcript ID opens the transcript page, which carries its own sections: an **Exons** table listing every exon with coordinates and phase, a **Sequence** view offering cDNA, CDS (coding sequence), and [protein sequence](/blog/guides/protein-sequence) with a translate toggle, and an **Exon/intron graphic** at higher resolution. From here you can copy a coding sequence straight into a primer design tool, a cloning plan, or a multiple alignment. A worked example. Suppose you need the coding sequence of the canonical CFTR transcript for a mutation nomenclature check. Search for CFTR, open the Transcripts tab, identify the row flagged Ensembl canonical (and, for human, MANE Select), click through to the transcript page, open the Sequence view, select CDS, and copy. The transcript you get is 27 exons long and encodes a protein of 1,480 amino acids, and every one of those facts is stated on the page without leaving the browser. ## The Variation Section The **Variation** tab lists sequence variants that Ensembl has mapped into the gene, drawn from dbSNP identifiers of the form rs plus a number. The table gives each variant's location, its alleles, and its most severe predicted consequence, using Ensembl's standardized consequence vocabulary: missense_variant, synonymous_variant, stop_gained, frameshift_variant, splice_donor_variant, intron_variant, and so on. Three uses of this section recur in coursework and analysis. **Filtering by consequence.** The table can be filtered to show only coding or only splice-region variants, which is how you quickly build a shortlist before running a full effect-prediction workflow with VEP, the Variant Effect Predictor, Ensembl's command-line and web tool for annotating variants with consequences, HGVS nomenclature, and scores. **Following through to variant detail.** Clicking a variant opens the variant page, which for human genes shows allele frequencies from the 1000 Genomes Project and other population resources, plus clinical annotations where sources such as ClinVar provide them. **Linking variation to phenotype.** Variants flagged in the phenotype tables (next section) often appear here with their disease association, giving you a bridge from molecular consequence to clinical relevance within two clicks. Keep the boundaries straight in your head: the gene page tells you which variants exist in the locus and what type of consequence each has. It does not, by itself, tell you whether a variant is pathogenic. Pathogenicity assessment requires the external clinical resources the page links to. ## The Phenotypes Section The **Phenotypes** tab collects associations between the gene and measurable traits or diseases, aggregated from sources including OMIM (Online [Mendelian Inheritance](/knowledge/molecular-biology/mendel-heredity-laws-of-inheritance-explained-simply) in Man) for human Mendelian disease, the GWAS Catalog for statistical trait associations, and, for mouse genes, the mammalian phenotype annotations maintained through mouse genetics resources. The table lists the phenotype, the source, and the supporting variant or study. This section is a rapid triage tool. Opening BRCA1 immediately surfaces its associations with hereditary breast and ovarian cancer through OMIM entries. Opening the mouse homolog of a human developmental gene surfaces the knockout phenotypes recorded for that mouse line. When you need to know whether a candidate gene from an expression experiment has any prior phenotype literature, this tab answers the question in ten seconds. The subtlety worth memorizing is the difference in evidence type. An OMIM association reflects accumulated clinical and familial evidence for a disease gene. A GWAS Catalog entry reflects a statistical association from a specific study, often for a complex trait, and often for a variant that may not even be causal. The tab presents both, and part of reading it correctly is noticing which source each row came from. ## Gene Expression The **Expression** tab shows where and how strongly the gene is transcribed, using baseline tissue data. For human genes, Ensembl presents expression summaries built from large public resources, with tissue-level panels such as those derived from GTEx (Genotype-Tissue Expression) data, and single-cell expression summaries in species where those datasets have been processed into the browser. Read the tab in two passes. First, the coarse question: is this gene restricted, preferential, or ubiquitous? A gene expressed almost exclusively in a single tissue is immediately more interesting as a candidate for a tissue-specific phenotype. Second, the magnitude question: how high is the expression relative to housekeeping controls in the same panel? The plots are normalized within the source dataset, so comparisons across genes are reasonable while comparisons across different source panels are not. A frequent student error is treating absence of expression in a panel as absence of expression anywhere. Expression panels cover defined tissue sets under defined conditions. Low or absent signal in a bulk tissue panel is compatible with strong expression in a rare cell type, which is precisely what single-cell views are there to reveal. ## Gene Tree and Homology: Comparative Relevance The **Gene tree** tab displays a phylogenetic tree, computed by Ensembl Compara, showing where the gene's protein family split across species. The **Homology** tab (labeled Orthologues in the interface) lists the counterpart genes in other species with their stable IDs, percent protein identity, and a confidence classification such as "One to one orthologue" or "Many to many." This is the section that makes Ensembl the standard first stop for comparative work. Two scenarios illustrate it. **Translating a finding across species.** You find that a gene is dysregulated in a mouse model and want to know the human counterpart. Open the mouse gene page, check Homology, and read the human orthologue row with its identity percentage. Ensembl classifies orthologues between species with well-annotated genomes with high confidence, and flags lower-confidence calls, for example between distant vertebrate lineages, so you know which equivalences are safe to build on. **Understanding a gene family.** The paralogues listed in the same tab show duplicates within the [human genome](/blog/guides/human-genome). For a transporter or receptor family, the paralogue list is the fastest way to see how many close relatives exist, which matters when interpreting drug specificity or knockout phenotypes, because a phenotype may be buffered by a paralogue. The tree graphic itself rewards careful reading. Branch lengths represent sequence divergence, and the tree typically spans a wide taxonomic range, from primates through mammals, vertebrates, and outgroups. For teaching molecular evolution, walking a class through the HBB gene tree and its globin-family paralogues is one of the clearest demonstrations that paralogue expansion and tissue-specific function go hand in hand. ## Regulation Where regulatory annotation exists for the species, the **Regulation** tab shows regulatory features overlapping the gene: regions of open chromatin, [histone modification](/knowledge/molecular-biology/histone-modification) peaks, and predicted [transcription factor](/knowledge/molecular-biology/transcription-factor) binding motifs, largely derived from the ENCODE project for human and corresponding resources for other species. The view is filtered by cell type, because a regulatory region active in a B cell line is not necessarily active in hepatocytes. For most gene-centered work, this tab is a checkpoint rather than a destination. Its practical value is checking whether an intronic variant from the variation table sits inside a regulatory feature in a relevant cell type, which upgrades the variant from "intronic, probably nothing" to "intronic, inside a promoter marked in the right tissue." ## Literature and External References Two final sections close the page. The **Literature** tab lists publications tied to the gene, retrieved through Europe PMC, ordered so that the most prominent papers surface first. It is a faster starting point than a raw PubMed query because the list is already filtered to the specific gene entity. **External references** (sometimes rendered as "External resources" in the interface) lists cross-links: RefSeq gene and transcript IDs at NCBI, UniProt protein entries, the HGNC record, OMIM entries, and [protein structure](/blog/guides/protein-structure) resources where applicable. This table is the answer to the perennial registry problem of knowing a gene in one database and needing its identity in every other. Copying the RefSeq and UniProt IDs from here, rather than searching each database by symbol, eliminates an entire class of wrong-gene errors caused by outdated symbols. ## How to Use Ensembl: A Practical Walkthrough ![Six numbered steps from searching Ensembl to branching out to variation, phenotypes, homology or expression tabs.](/article-images/illustrations/ensembl-gene-page-sections-gene-page-walkthrough.webp) This section functions as a compact Ensembl [genome browser tutorial](/blog/guides/genome-browsers-for-researchers-a-guide-to-inspecting-genomic-evidence) built around the gene page. Follow it once with a real gene and the interface stops being intimidating. **Step 1. Search.** Type the gene symbol, an Ensembl ID, a plain-text description, or even raw chromosomal coordinates into the search box at the top of ensembl.org. Search accepts "BRCA1," "ENSG00000012048," or "17:43044295-43125482." Ambiguities are surfaced immediately: searching a symbol that matches several species or several loci returns a disambiguation page listing every hit with species, location, and description. **Step 2. Confirm species and assembly.** The page header states the species and, in the location strip, the assembly. If your project is built on GRCh37, use the dedicated GRCh37 archive site rather than trying to mentally convert coordinates. **Step 3. Read the header.** Confirm the biotype and the stable ID version. If you are about to cite this gene in a methods section, record the stable ID with version and the release number now. **Step 4. Open Transcripts.** Identify the canonical and, for human, the MANE Select transcript. Note the exon count and protein length. Click through to the transcript page if you need cDNA, CDS, or protein sequence. **Step 5. Take what you need.** Use Download GTF in the header for annotation, the transcript page for sequences, or the export dialog for custom formats. **Step 6. Branch out by question.** Variant questions go to Variation, disease questions to Phenotypes, cross-species questions to Homology, tissue questions to Expression. ### Getting Data Out Programmatically The gene page is the human-readable front door to programmatic services that mirror it. - **BioMart** (via the Tools menu) lets you build a table query, for example all human protein-coding genes on chromosome 17 with their coordinates and RefSeq matches, and export it as TSV. - **The Ensembl REST API** reproduces every section over HTTP. A request to `https://rest.ensembl.org/lookup/id/ENSG00000012048?content-type=application/json` returns the gene's coordinates, biotype, and version as JSON. Endpoint families exist for sequence retrieval, homology, variation, and overlap queries. - **The FTP site** (ftp.ensembl.org) distributes the full GTF, GFF3, FASTA, and variant files for every species and release, which is what large pipelines actually consume. The workflow that scales is: prototype on the gene page interactively, then reproduce the same query in BioMart or REST once you know exactly which fields you need. ### Versioning and the Archive Ensembl publishes several releases per year, and a release number is part of any reproducible claim. When you need the annotation exactly as it existed in an earlier release, for example because a published paper analyzed release 75, use **archive.ensembl.org**, which hosts snapshots of past releases, and the **ID History Converter** tool, which maps an old stable ID to its current equivalent. These two resources resolve the majority of "the gene page looks different from the paper" situations. ## How the Gene Page Maps to Other Genome Browsers Students arriving from NCBI or UCSC need a translation table more than a new vocabulary. The **NCBI Gene record** is the closest analogue to the Ensembl gene page summary, but NCBI's sequence browsing lives on separate RefSeq pages, whereas Ensembl integrates annotation, variation, and comparative data under one gene entity. The gene models themselves differ by design: Ensembl and RefSeq annotate transcripts independently, which is why the same human gene can show a slightly different exon count in each, and why the MANE collaboration exists to nominate a shared transcript per gene. The **[UCSC Genome Browser](/knowledge/bioinformatics/ucsc-genome-browser-beginner-tutorial-and-tips)** is organized around the coordinate view rather than the gene entity. Its strength is dense third-party track collections. Ensembl's gene page is stronger as a structured summary per gene and as a source of Ensembl's own consistent annotation across hundreds of species. **UniProt** operates at the protein level. Where an Ensembl transcript page gives you a protein sequence and its coding coordinates, a UniProt entry gives you curated functional annotation, domains, and literature-based evidence. The external references section of the gene page is the link between the two. ## Common Mistakes and Limitations These are the specific errors that show up most often in teaching and review. **Quoting coordinates without the assembly.** "TP53 is at chr17:7,668,421-7,687,550" is incomplete information. Every coordinate belongs to a named assembly, and GRCh37 and GRCh38 positions for the same variant differ. Always record assembly, release, and stable ID version together. **Treating the Ensembl canonical transcript as an biological truth.** The canonical flag is an annotation policy choice, not a statement that other transcripts are artifacts. For human clinical work, prefer the MANE Select transcript, which both Ensembl and NCBI agree on, and say explicitly which transcript any HGVS nomenclature refers to. **Assuming the gene ID never changes.** The core stable ID is stable across releases, but the version suffix changes when annotation changes, and rare cases of merged or retired IDs do occur. For longitudinal studies, snapshot your gene list per release and use the ID History Converter when reconciling across releases. **Confusing Ensembl with GENCODE.** For human and mouse, Ensembl and GENCODE share the same underlying annotation pipeline and gene models, differing mainly in release schedule and in the set of transcript-level flags such as GENCODE basic. Reporting "Ensembl release 110 annotation" and "GENCODE release 44" refers to closely matched but not literally identical file sets. **Reading the transcript table as a list of genes.** A gene page can list dozens of transcripts. The gene is the locus, and the transcripts are alternative versions of it. Analyses that count "genes" by counting transcript rows will inflate their numbers severalfold. **Over-reading the phenotype table.** A GWAS Catalog row is one study's statistical association, not settled causation, and an OMIM entry aggregates curated evidence of varying depth. Check the source column before repeating an association. **Using the browser where the API is the right tool.** Clicking through hundreds of gene pages is not reproducible. Once a query is defined, move it to BioMart, REST, or the FTP files so it can be scripted and re-run. **Expecting every section in every species.** Regulation tracks, deep expression data, and rich phenotype tables are strongest for well-resourced species such as human, mouse, rat, and zebrafish. For a species sequenced primarily for comparative purposes, several tabs will be sparse, and that reflects the underlying data, not a browser malfunction. Finally, a note on scope: this guide explains how the browser is organized and how to read it. Scientific judgment about what a particular annotation, variant, or phenotype association means for a specific organism, patient, or breeding program belongs with the relevant domain expert and the primary literature for that question. ## Quick Review - The gene page is organized as a header block plus tabs: Location, Transcripts, Variation, Phenotypes, Expression, Gene tree and Homology, Regulation, Literature, and External references. - Every gene carries a stable ID (for example ENSG00000012048) with a version suffix, and reproducibility requires recording ID, version, assembly, and release number together. - The Transcripts tab is the analytical core: it lists all transcripts with biotype, protein length, and flags including Ensembl canonical, MANE Select, CCDS, and TSL. - Coordinates are assembly-specific. Human annotation sits on GRCh38 and mouse on GRCm39, and positions are not comparable across assemblies. - The Variation tab lists variants with standardized consequence types, and links to VEP for full effect prediction and to variant pages for frequencies. - The Homology tab provides orthologues and paralogues with identity and confidence classes, computed by Ensembl Compara. - For bulk or programmatic work, take the same data from the Download GTF button, BioMart, the REST API, or the FTP site rather than clicking page by page. ## Frequently Asked Questions ### What are the Ensembl gene page sections? They are the labeled blocks and tabs that organize the information Ensembl holds for a single gene: a summary header with the stable ID and genomic location, followed by tabs for transcripts, variation, phenotypes, expression, homology, regulation, literature, and external cross-references. ### Which transcript on the Ensembl gene page should I use? ![Decision path for choosing MANE Select, Ensembl canonical or another transcript, and recording the choice with assembly and release.](/article-images/illustrations/ensembl-gene-page-sections-transcript-choice.webp) For human protein-coding genes, use the transcript flagged MANE Select when one exists, because it is the jointly agreed Ensembl and RefSeq representative intended for clinical reporting. Otherwise the Ensembl canonical flag marks Ensembl's recommended default, and you should always state which transcript your analysis used. ### How do I download a gene sequence and its annotation from Ensembl? On the gene page, use the Download GTF button for annotation and the transcript page's Sequence view for cDNA, CDS, and protein sequences in FASTA. For many genes at once, use BioMart, the REST API, or the FTP site files for the relevant release. ### Why did the Ensembl gene ID or coordinates change between releases? Stable IDs persist across releases but gain a higher version suffix when their annotation is revised, and coordinates can shift between reference assemblies such as GRCh37 and GRCh38. Use archive.ensembl.org for past releases and the ID History Converter to map old IDs to current ones. ### How do I find the mouse or other species counterpart of a human gene? Open the human gene page, go to the Homology (Orthologues) tab, and read the orthologue table, which lists counterpart genes with their stable IDs, percent identity, and a confidence classification from Ensembl Compara. ### Can I use Ensembl for species other than human? Yes. Ensembl maintains annotation for hundreds of vertebrate species in the main browser, with separate divisions for plants, animals beyond vertebrates, fungi, protists, and bacteria, and the same gene page structure applies, though data-rich tabs such as Regulation and Expression are fuller for well-studied species.