# Ensembl Browser Navigation Tips By the end of this guide you will be able to find any gene in Ensembl, move confidently between the Location, Gene, Transcript, and Sequence views, configure tracks so the display shows what you actually need, zoom and pan without losing your place, export sequences and annotation in standard formats, upload your own variants as a custom track, and pull the same data from the command line through the Ensembl REST API. You will also know the handful of mistakes that waste the most time, especially the assembly mismatch trap. What you need on hand: - A modern web browser. Nothing to install. The browser lives at ensembl.org. - A gene name, variant identifier, or genomic coordinate you want to look at. If you have nothing in mind, use BRCA2 as your practice gene. - Optional but useful: a terminal with curl installed if you want the command line companion section, and a VCF file if you want to upload your own variants. This guide covers how to use Ensembl as a working tool. It is written as a practical ensembl [genome browser tutorial](/blog/guides/genome-browsers-for-researchers-a-guide-to-inspecting-genomic-evidence) in walking pace, with each step giving a real action, an expected result, and a way to check that you got it right. ## What the Ensembl Browser Actually Is A genome browser is a scrollable map of a genome. Ensembl is one of the major public genome browsers, jointly produced by EMBL's [European Bioinformatics Institute](/knowledge/bioinformatics/the-european-bioinformatics-institute-embl-ebi) and the Wellcome Sanger Institute. It holds annotated reference genomes for vertebrates and a wide range of other species, along with genes, transcripts, regulatory features, variants, [comparative genomics](/blog/guides/comparative-genomics) data, and cross references to databases such as UniProt and NCBI. Three vocabulary terms will follow you through the whole article: - **Assembly.** The reference genome sequence itself, for example GRCh38, the current human reference assembly. Coordinates only mean something within a specific assembly. - **Track.** A horizontal data layer drawn against the genome, for example a track of genes, a track of repeat regions, or a track of your own uploaded variants. - **Coordinate.** A position on the assembly, written as chromosome, start, end, for example 13:32315474-32400266. Ensembl uses 1-based coordinates that include both endpoints. Ensembl is also deeply embedded in the wider bioinformatics ecosystem. Specialist model organism databases build their own portals and link out to Ensembl for the underlying genome. Xenbase, the Xenopus database, maintains its own genome browsing through GBrowse and cross links gene pages to Ensembl alongside NCBI and UniProt, so researchers routinely arrive at Ensembl from a model organism site and need to orient themselves quickly [1]. Conversely, projects that assemble individual or non-reference genomes frequently present their data in the context of Ensembl annotations. The HuRef browser, built around the genome of a single individual, provided a comparative view between the NCBI human reference and that individual's assembly while displaying features in the context of existing Ensembl annotations, an early demonstration of how valuable an Ensembl-anchored frame of reference is when you are looking at anything other than the bare reference [2]. Keep that frame of reference idea in mind. Most navigation confusion in Ensembl comes from not knowing which genome, which assembly, and which view you are currently standing in. ## Getting Started: Find Your First Gene in Five Minutes ![Five step flow from the ensembl.org front page to the Location view for human BRCA2 on GRCh38.](/article-images/illustrations/ensembl-browser-navigation-tips-first-gene-route.webp) This is the fastest possible ensembl browser navigation tips step by step route from zero to a meaningful view. **Step 1. Open the site and pick your species.** Go to ensembl.org. The front page lists popular species with their assembly names, for example Homo sapiens GRCh38 and Mus musculus GRCm39. Click your species name, or type the species into the search box. **Step 2. Search by gene symbol.** Type a gene symbol such as BRCA2 into the main search box and press Enter. You will land on a search results page grouped by category, with genes listed first. Each gene result shows the symbol, a descriptive name, the species, and the location on the assembly. **How to check you got it right.** The gene result for human BRCA2 should report chromosome 13 and a start position in the region of 32.3 million base pairs on GRCh38. If you see a different chromosome, you either searched the wrong species or hit a gene name synonym from another organism. **Step 3. Open the gene page.** Click the gene name. You land on the Gene tab, which shows a summary card with the Ensembl gene ID, for example ENSG00000139618 for human BRCA2, the biotype such as protein coding, the strand, and a compact graphic of the transcripts. **Step 4. Notice the URL.** Look at the address bar. It contains the species, the view type, and the identifier, in a pattern like this. ``` https://www.ensembl.org/Homo_sapiens/Gene/Summary?g=ENSG00000139618 ``` Every Ensembl view has a bookmarkable URL of this shape. Learning to read and edit these URLs is one of the highest value navigation skills, covered in detail below. **Step 5. Jump to the genomic context.** In the left hand menu of the Gene tab, click "Location". Now you are in the Location view, the scrollable map. This is where most real navigation work happens, so the next sections break it down. ## The Core Views: Location, Gene, Transcript, and Sequence Ensembl organizes information into tabs. Each tab is a different answer to a different question about the same genomic feature. ### The Location tab, the map itself The Location tab answers "what is at this position in the genome?" It draws the chromosome, a zoomed region, and a basepair view, all stacked vertically and all linked. Three panels matter most. - **Chromosome panel.** A whole chromosome ideogram with a red box marking where you are. Click anywhere on it to jump. - **Region in detail.** The main working view. Genes above and below the line are transcribed from opposite strands. Click any gene, transcript, variant, or regulatory feature for a popup with details and links. - **Basepair view.** Appears when you zoom below about 1 kilobase wide, showing actual sequence letters and, where relevant, translated codons. Navigation gestures in the region view: - **Drag to pan.** Click and drag the main panel horizontally to slide along the chromosome. The panels stay in sync. - **Drag to select, then zoom.** Drag a box over a region and choose "Zoom to selection" or right drag options to jump to just that interval. - **Click the ruler.** Click a position on the coordinate ruler to center the view there at the current zoom level. - **Zoom buttons and the scale bar.** Use the plus and minus zoom controls or type an exact width in bases. - **Flip the strand.** The "Reverse" control flips the display so genes on the minus strand read left to right. This affects display only, never the underlying coordinates, which is a frequent source of confusion covered under common mistakes. ### The Gene tab, the annotation record The Gene tab answers "what is this gene?" It lists the gene ID, biotype, synonyms, external references such as HGNC symbols and UniProt accessions, a transcript table, and links to phenotype and variation data. The transcript table is worth slow reading. Each row gives a transcript ID such as ENST00000544455, whether that transcript is flagged as a canonical transcript, whether it has RefSeq or UniProt protein support, and the length of the protein it encodes. ### The Transcript tab, one isoform at a time Click any transcript ID in the gene page. The Transcript tab shows the exon structure, the coding sequence boundaries, the protein product, and domains. This is the view to use when someone gives you a variant described relative to a specific transcript, for example a c. position from a clinical report. Confirm the transcript ID on the report matches the transcript you are viewing, otherwise the c. position will not map where you expect. ### The Sequence tab, plain text you can copy The Sequence tab shows the genomic, cDNA, or [protein sequence](/blog/guides/protein-sequence) with features highlighted in context. Use the "Show sequence" controls to switch between genomic DNA, cDNA, and protein, and to toggle the display of translation, variants, and regulatory marks overlaid on the sequence. For anything longer than a screenful, skip the display and export instead, covered below. ### The Variation and Phenotype tabs If the gene has known variants, the Variation tab lists them with consequence type, allele frequencies where population data exists, and clinical significance. The Phenotype tab lists trait and disease associations with their supporting publications. These tabs are gateways to the variant level views, which share the same navigation grammar as everything else. ## Ensembl Browser Navigation Tips That Save the Most Time These are the working tips that separate a frustrating session from a productive one. Each one is a real click path you can follow today. ### Tip 1. Treat URLs as a command language Ensembl URLs encode state, which means you can navigate by typing them. The most useful pattern is the region URL. ``` https://www.ensembl.org/Homo_sapiens/Location/View?r=13:32315474-32400266 ``` The `r` parameter is the region. Edit the chromosome and coordinates directly in the address bar, press Enter, and the browser jumps there. Useful variations: - Add `;db=core` style parameters only if you know you need them. For everyday work, `r` plus the view name is enough. - To center a coordinate at a given width, compute the window yourself. To center on position 32330000 with a 20 kilobase window, set the region to 13:32320000-32340000. The arithmetic is start = center minus half the width, end = center plus half the width. - Bookmark or share the URL after you configure tracks and zoom. The URL preserves the region, though track configuration is tied to your account or browser storage, so share URLs with the expectation that the recipient sees default tracks unless they configure their own. ### Tip 2. Configure tracks deliberately, not by default The default track set is a compromise. On the Location view, click "Configure this page" in the left menu. You get a catalog of track categories, including genes, variation, regulation, comparative genomics, and your own custom data. Practical settings: - Turn off tracks you do not use. Every active track costs drawing time, and on some species the default stack of predictions and alignments can crowd out the annotation you care about. - For cross species work, turn on the conservation or pairwise alignment tracks for your species of interest. These render as blocks of sequence similarity in the comparative genomics section. - For human variant work, enable the variation tracks and set density to a level you can read. At wide zoom, dense variation tracks become visual noise. Zoom in before judging them. - Click the checkmark icon to save. Configuration persists for your session and, if you log in to a free Ensembl account, across sessions. ### Tip 3. Distinguish the two configuration menus "Configure this page" changes track rendering for the view you are in. "Configure region" in the region in detail panel changes what that specific panel draws. If a track refuses to appear, check both. This is one of the most common ensembl browser navigation tips tips that experienced users tell beginners, because the two menus look almost identical. ### Tip 4. Use the search box as a universal translator The search box accepts gene symbols, Ensembl IDs, dbSNP reference SNP identifiers such as rs numbers, protein accessions, and plain coordinates in the form 13:32315474-32400266. Typing a coordinate directly is often faster than any menu. If the search box offers you an autocomplete list, choose the entry that names the species and assembly you expect. ### Tip 5. Know your assembly before you type a single coordinate Human data in Ensembl lives on GRCh38 by default. A large body of older literature, older annotation files, and some clinical pipelines still use GRCh37 coordinates. Ensembl maintains a dedicated archive of the browser on the GRCh37 assembly so you can look up old coordinates without converting them. If a coordinate lookup returns nothing or a completely different gene than you expected, an assembly mismatch is the first thing to suspect. This is the single most expensive mistake in genome browsing, and it has its own section below. ### Tip 6. Export instead of screenshot The left menu of most views includes an "Export data" option. From it you can pull the current region as plain sequence in FASTA format, or the annotation as GFF or other tabular formats, for exactly the coordinates you are viewing. Screenshots lose the underlying data. Exports preserve it. For whole gene or whole genome scale retrieval, use BioMart or the download files at the Ensembl FTP site rather than clicking region by region. ### Tip 7. Use the strand controls and read the arrows Genes above the horizontal line in the region view run in one direction and genes below run in the other. Transcript arrows point in the direction of transcription. If a sequence you exported reads backward relative to your expectation, you either exported from the reverse strand or copied the reverse complement by mistake. The Sequence view includes explicit controls for reverse complementing the display. ### Tip 8. Upload your own data as a custom track Under "Add your data" in the left menu, you can attach your own variants or intervals. Ensembl accepts standard formats including VCF for variants and BED style interval files, either uploaded as a file or attached by URL. Your data then renders as a track alongside the reference annotation, which is the fastest way to eyeball where your calls land relative to genes. This is also where the individual genome perspective matters. Reference browsers show one haploid consensus sequence, while real diploid individuals carry two haplotypes that differ from it. Projects that display an individual's assembly alongside a reference, such as the HuRef browser, made this explicit by rendering single nucleotide variants, indels, inversions, and copy number differences in the context of reference annotations, and by giving access to the underlying sequence reads and quality values supporting each call [2]. When you upload a VCF to Ensembl and see your heterozygous sites painted over the reference genes, you are doing a lightweight version of the same exercise. ### Tip 9. Use BLAST and BLAT inside the browser The Tools menu at the top of any Ensembl page includes BLAST and BLAT. Paste a raw sequence and Ensembl will locate matching regions in the species you choose. This rescues the common situation where you have a sequence fragment and no identifier. Run BLAST against the species you expect, open the hit, and you are dropped straight into the Location view at the match. ### Tip 10. Bookmark species and build a personal species bar After you log in to a free Ensembl account, you can save frequently used species so they appear as one click links at the top of the page. If you work across human, mouse, and zebrafish daily, this removes hundreds of clicks per week. ## Worked Walkthrough: From a Variant rsID to a Protein Change Here is a complete ensembl browser navigation tips step by step example that exercises most of the skills above. The task is to find out what a coding variant does to the protein. **Step 1. Search the variant.** In the Ensembl search box, type a dbSNP identifier such as rs28897772 and press Enter. The results page shows the variant across relevant species. Choose the human entry. **Step 2. Land on the Variation tab.** The page reports the variant's location on GRCh38, its allele pair, and its consequence type, for example missense variant. The consequence shown here is a summary. Note the transcript it is calculated against, listed by Ensembl transcript ID. **Expected output check.** You should see a coordinate of the form 13:xxxxxxx on chromosome 13, an allele string such as A/G, and a consequence classification. If the coordinate is not on the chromosome you expect, stop and resolve the assembly or identifier mismatch before continuing. **Step 3. Open the genomic context.** Click "Location" in the left menu. The region view opens centered on the variant, with the variant drawn as a marker on the sequence and the gene model drawn around it. **Step 4. Inspect the transcript context.** Click the variant marker, then in the popup follow the link to the transcript whose consequence you care about. On the Transcript view, find the exon containing the variant and check its position within the coding sequence. **Step 5. View the codon change.** Zoom the region view down until the basepair panel appears. With the variant track enabled and translation enabled in the sequence display, you can see the reference codon, the variant base within it, and the translated amino acids before and after. **Step 6. Get the formal consequence prediction.** In the variation page or via the transcript context menu, open the consequence details. For missense variants, Ensembl reports the amino acid change in the standard three letter notation, for example p.(Ala2486Gly) style output, along with any predicted impact scores from tools such as SIFT where available. **Step 7. Export the evidence.** Use "Export data" on the region view to save the surrounding sequence, and use the export options on the variation page for the variant record itself. You now have a text record of what you saw, which belongs in your [lab notebook](/knowledge/molecular-biology/lab-notebook) or supplementary material. **Step 8. Record the exact transcript version.** Note the full transcript ID from the page. Any future reanalysis should use the same transcript, because amino acid numbering is transcript dependent. ## Command Line Companion: The Ensembl REST API Everything you can see in the browser is also available programmatically through the Ensembl REST API at rest.ensembl.org. This matters when a colleague asks you to check forty variants, not one. ### Example 1. Look up a gene and get its canonical transcript ```bash curl -H "Content-Type: application/json" \ "https://rest.ensembl.org/lookup/id/ENSG00000139618?expand=1" ``` The response is a JSON object describing the gene, including its assembly coordinates and, because of the expand parameter, a list of its transcripts. Each transcript entry carries an `is_canonical` field. Grep for it if you only want the canonical transcript. ```bash curl -s -H "Content-Type: application/json" \ "https://rest.ensembl.org/lookup/id/ENSG00000139618?expand=1" \ | python3 -c "import json,sys; d=json.load(sys.stdin); \ print([t['id'] for t in d['Transcript'] if t.get('is_canonical')==1])" ``` Expected output for BRCA2 is a single transcript ID, the canonical protein coding transcript. ### Example 2. Fetch the sequence of a region ```bash curl -H "Content-Type: text/plain" \ "https://rest.ensembl.org/sequence/region/human/13:32315474-32400266" ``` Expected output is a plain FASTA record with a header naming the region and 84,793 bases of sequence. The base count equals end minus start plus one, which is the 1-based inclusive convention Ensembl uses everywhere. Use that equality as a sanity check on any region you fetch. ### Example 3. Look up a variant ```bash curl -H "Content-Type: application/json" \ "https://rest.ensembl.org/vep/human/rs28897772" ``` This asks the Variant Effect Predictor endpoint to annotate the variant. The response lists each transcript the variant overlaps, with the consequence term, the codon and amino acid change where applicable, and population frequency data where the species has it. This is the programmatic version of steps 1 through 6 in the walkthrough above. Rate limits apply to the public API, so batch requests politely and cache results locally rather than re-querying the same identifier in a loop. For heavy work, install the command line Variant Effect Predictor and run it against a local cache. ### Parameters that matter in the walkthrough commands | Parameter or control | Where it applies | What it does | Practical note | |--|--|--|--| | `r=chr:start-end` | Browser URL | Jumps to an exact region | Compute the window yourself around a center point | | `expand=1` | REST lookup | Includes child features, for example transcripts of a gene | Set `expand=0` or omit for a smaller response | | `Content-Type: application/json` | REST header | Returns structured JSON | Use `text/plain` when you want raw sequence | | Configure this page | Location view | Turns track categories on and off | Fewer active tracks means faster pages | | Reverse | Location view | Flips display strand | Changes nothing in the stored coordinates | | Export data | Most views | Saves FASTA, GFF, or [tabular data](/blog/research-skills/tabular-data-what-it-is-and-how-to-analyze-it) | Always prefer this over screenshots | | Add your data | Left menu | Uploads VCF or interval files as custom tracks | Host by URL for large files | | GRCh37 archive site | Human browsing | Serves the previous assembly | Use for legacy coordinates instead of converting by hand | ## Ensembl in the Wider Ecosystem: Model Organisms and Comparative Browsing Two patterns from the published literature describe how Ensembl fits into broader workflows, and both have direct navigation consequences. The first pattern is the specialist model organism portal that sits in front of Ensembl-style data. Xenbase, the Xenopus model organism database, hosts current and archived genomes for both Xenopus laevis and Xenopus tropicalis, provides BLAST and GBrowse browsing on its own site, curates [gene expression](/blog/guides/gene-expression) images and gene ontology annotations onto gene specific pages, and links each gene out to Ensembl, NCBI Entrez, and UniProtKB [1]. The navigation lesson is that the organism database is often the better starting point for biology questions such as expression patterns, reagents, and literature, while Ensembl is the better tool for genome arithmetic, transcript structure, and comparative genomics. Learn to move between them through the cross links rather than fighting either interface to do a job it was not built for. The second pattern is comparative browsing against a reference. The HuRef browser demonstrated a comparative view between the standard human reference and the assembled genome of a single individual, letting users see SNPs, indels, inversions, and structural and copy number variants in the context of functional annotations on either genome, with full access to the underlying reads and their quality values [2]. Modern Ensembl supports the same spirit of comparison in two ways, through pairwise and multiple alignment tracks between species, and through user uploaded custom data tracks that overlay your variants or intervals on the reference. Both rely on the coordinate discipline described above. Alignment tracks are only interpretable when you know which assembly each panel is drawn on, which Ensembl labels on screen, and custom data uploads only land in the right place when their coordinates match the current assembly. A practical comparative navigation sequence: 1. Open your gene on the Location view in your primary species. 2. Open "Configure this page", go to the comparative genomics section, and enable conservation or alignment tracks for your target species. 3. Identify the conserved block containing your region. Click through the alignment to see the equivalent region in the other species. 4. If the other species has a specialist database, follow the cross links from its Ensembl gene page, in the same way Xenbase gene pages link outward to Ensembl and other major resources [1]. ## Common Mistakes and Limitations ![Decision path for a coordinate that returns nothing or the wrong gene, branching on assembly, strand and identifier type.](/article-images/illustrations/ensembl-browser-navigation-tips-coordinate-triage.webp) **Mixing up GRCh37 and GRCh38 coordinates.** This is the costliest error. A coordinate that points into BRCA2 on GRCh38 can point somewhere unrelated on GRCh37, and vice versa. Every coordinate must be paired with an assembly name in your notes, scripts, and publications. When you inherit a coordinate from a colleague or an old paper, ask which assembly it came from before you search for it. **Assuming coordinates transfer between assemblies.** Within an assembly, coordinates are stable. Between assemblies, they are not, because the reference sequence itself changed. You need a liftover step or a cross assembly mapping, not arithmetic. Ensembl's archive sites let you browse legacy assemblies directly, which is often simpler than converting coordinates. **Quoting an amino acid change without naming the transcript.** The same genomic variant can produce different c. and p. descriptions on different transcripts. Clinical and publications standards require the transcript reference. Check the transcript ID on the Ensembl page against the one in your source document. **Confusing Ensembl IDs with external IDs.** ENSG numbers are genes, ENST numbers are transcripts, ENSP numbers are proteins, and the digits change when annotation is revised. A gene ID from a years old dataset may have been merged or retired. The Gene tab lists current external cross references, and the search box will resolve many retired IDs to their current entry. **Reading the display strand instead of the annotation strand.** Flipping the display with the reverse control changes which way genes are drawn on your screen and nothing else. Exported sequences and reported coordinates are unaffected. If a sequence looks reversed, check which strand you exported from before assuming the data is wrong. **Judging variation density at the wrong zoom level.** At whole chromosome zoom, variation tracks render as nearly solid bars. That is a rendering artifact of density, not evidence that a region is saturated with variants. Zoom in before interpreting. **Treating the reference as a real individual genome.** The reference assembly is a consensus artifact. Individuals carry two haplotypes that both differ from it in SNPs, indels, and structural variation, and projects such as HuRef made those differences visible against the annotations in a diploid context [2]. Interpret your own variant calls in that light. **Losing work in an unconfigured default view.** If you configure tracks carefully and then clear cookies or move to another machine without logging in, your configuration goes with the session. Create the free Ensembl account and save your species list and settings. **Retrieving large datasets one page at a time.** Clicking region by region for hundreds of genes is a workflow failure. Use BioMart for structured queries, the FTP site for bulk annotation files, and the REST API for scripted access. The browser is for looking, not for bulk export. **Forgetting that consequence predictions are predictions.** Consequence terms and pathogenicity scores on variation pages are computational annotations with defined limitations, not clinical diagnoses. Individual cases involving health decisions need qualified human genetics professionals, in the same way that animal cases need a veterinarian. ## Frequently Asked Questions ### How do I search for a gene in Ensembl? Type the gene symbol, Ensembl ID, or full name into the search box at the top of ensembl.org, then choose the gene entry for the correct species from the results. If the symbol is ambiguous, the results page lists all matches grouped by species and feature type. ### What is the difference between GRCh37 and GRCh38 in Ensembl? GRCh38 is the current human reference assembly that Ensembl serves by default, while GRCh37 is the previous assembly that remains available on Ensembl's dedicated archive site for legacy coordinates. Coordinates differ between the two, so always record which assembly your data uses. ### How do I download a sequence or annotation from the browser? Open the view you want, choose "Export data" in the left menu, and select FASTA for sequence or GFF and tabular formats for annotation over your current region. For genome scale downloads, use BioMart or the Ensembl FTP site. ### Can I upload my own data as a track in Ensembl? Yes. Use the "Add your data" option in the left menu to upload VCF, interval, or similar files, or point Ensembl at a hosted file by URL. Your data renders as a custom track alongside the reference annotation. ### How do I compare my gene across species in Ensembl? Open the gene, go to the Location view, enable comparative genomics or conservation tracks through "Configure this page", and click through the alignment blocks to the equivalent region in the other species. The orthologue relationships are also listed on the gene page. ### What is the Ensembl REST API and when should I use it? The REST API at rest.ensembl.org returns the browser's data as JSON or plain text over HTTP, which makes it the right tool when you need to check many identifiers in a script. Use the browser for visual exploration and the API for anything repetitive. ### Does Ensembl cover non-human and non-model organisms? Ensembl covers a broad set of vertebrates and additional species, and specialist organism databases such as Xenbase provide organism specific curation and link out to Ensembl for genome level data. For deep biology on a model organism, start at its organism database and cross link to Ensembl. ### Why does a coordinate I copied give a different gene than expected? The usual cause is an assembly mismatch between where the coordinate came from and the assembly currently shown, with strand confusion as the second most common cause. Confirm the assembly on both ends before converting anything, and prefer browsing the matching archive assembly over hand converting coordinates. ## Sources 1. [Navigating Xenbase: An Integrated Xenopus Genomics and Gene Expression Database.](https://pubmed.ncbi.nlm.nih.gov/29761462/) 2. [The HuRef Browser: a web resource for individual human genomics.](https://pubmed.ncbi.nlm.nih.gov/19036787/)