# Using NCBI Taxonomy to Explore Evolutionary Relationships

NCBI Taxonomy is a curated database maintained by the National Center for Biotechnology Information that places every organism with sequence or genome data in GenBank onto a single, shared classification tree. Using NCBI taxonomy to explore evolutionary relationships means reading that tree to trace lineages, connecting each node to the sequences behind it, and then testing those relationships with your own phylogenetic analysis.

This matters because almost every question in comparative biology starts with a species list and a question about relatedness. Which genera are close enough to compare across? Which virus lineage does this new sequence belong to? Which mammal should serve as the outgroup for a gene family analysis? NCBI Taxonomy answers the framing question before you ever open an alignment window, and because every taxonomic node links directly to GenBank, SRA, and genome assemblies, it doubles as the index for the largest collection of biological sequence data in the world. This article is a practical ncbi databases guide for that workflow, with real commands, interpretation principles, and the pitfalls that trip up new analysts.

## What Does NCBI Taxonomy Evolutionary Relationships Mean

When people search for "ncbi taxonomy evolutionary relationships explained," they usually want to know two things at once. First, what is stored in the database. Second, what the tree can and cannot tell you about evolution. Both deserve a direct answer.

NCBI Taxonomy stores, for each recognized organism, a unique numeric identifier called a **taxid**, one or more names (scientific name, synonyms, common names), a **rank** (species, genus, family, and so on), and a **lineage**, which is the ordered chain of ancestors from that organism up to the root of the tree. If you look up human (*Homo sapiens*), you get taxid 9606 and a lineage that runs through *Homo*, Hominidae, Primates, Mammalia, and upward through Eukaryota to the root. *Escherichia coli* sits at taxid 562 under Gammaproteobacteria. Every sequence record in GenBank carries one of these taxids, which is what makes the taxonomy database the [connective tissue](/knowledge/veterinary-medicine/veterinary-anatomy-physiology/connective-tissue-types-loose-dense-and-more) of NCBI.

So what does ncbi taxonomy evolutionary relationships mean in practice? It means the database encodes a **consensus classification** that reflects current scientific agreement about relatedness, but it is not itself a peer-reviewed phylogenetic analysis of every node. Nodes are added and reshaped as published evidence accumulates and as curators at NCBI and partner taxonomy resources review the literature. The tree is largely binary where relationships are well resolved and collapses into **polytomies** (a node with more than two immediate descendants) where the order of divergence is uncertain or contested. A polytomy is not an error. It is an honest signal that the data on file do not establish a single branching order.

The practical consequence is that NCBI Taxonomy is best treated as a map and an index rather than a final evolutionary answer. It tells you the accepted framework of names and groupings, gives you an entry point to every sequence associated with each group, and highlights where classification is unstable. To test the relationships themselves, you retrieve sequences through the taxonomy tree and build your own phylogeny, which is exactly what the published studies used as worked examples below do.

## How the NCBI Taxonomy Database Is Structured

### The core files and identifiers

The entire database ships as a small flat-file archive called **taxdump**, available from the NCBI FTP site (ftp.ncbi.nlm.nih.gov/pub/taxonomy/taxdump.tar.gz). The two files you will use most are:

1. **names.dmp**, which maps every taxid to its names, including the scientific name, synonyms, and common names.
2. **nodes.dmp**, which maps every taxid to its parent taxid, its rank, and additional flags such as whether the node has a GenBank hidden flag.

Two smaller files handle change management. **merged.dmp** lists taxids that were merged into other taxids, and **delnodes.dmp** lists taxids that were deleted outright. These files exist because taxids are not permanent. When two populations previously treated as separate species are synonymized, one taxid can be merged into another. Scripts that hard-code taxids without checking these files break quietly, which is a common cause of stale downstream analyses.

Above the species level, the backbone includes familiar ranks such as genus, family, order, class, phylum, and superkingdom (the top-level rank holding Bacteria, Archaea, and Eukaryota). Viruses occupy their own branch of the tree, and their internal structure has been progressively aligned with the formal virus classification used by the field. Many intermediate nodes carry no formal rank at all, which surprises newcomers. An unnamed node with no rank is simply a grouping the curators found useful for organizing organisms and linking data, and it should be interpreted as a clade without a rank label rather than as an error.

### The Taxonomy Browser and tree display

The public face of the database is the Taxonomy Browser at ncbi.nlm.nih.gov/taxonomy. Type a name such as "Rhaphiolepis" and you land on a page showing the complete lineage from the root down, the number of linked nucleotide, protein, and genome records, and the immediate children of the node. The browser offers several display modes, including a graphic tree view, a taxonomic blaster for retrieving sequences, and an Entrez search restricted to that node and its descendants. The "entire taxonomy" display mode walks the whole tree from the root, which is useful for seeing where your group of interest sits relative to everything else.

One detail worth memorizing: the browser shows **number of descendants** next to each node, and those counts are live. If a genus shows only two species with nucleotide records while a related genus shows hundreds, that is sampling reality in GenBank, not biology. Sampling depth varies enormously across the tree of life, and any comparison you make later depends on it.

### Names, synonyms, and the "Organism" field

Every Entrez query across NCBI accepts an organism limit using the bracketed field `[Organism]`. The query `Rosaceae[Organism]` in the Nucleotide database returns sequences from every species under the rose family in the taxonomy tree, including children, grandchildren, and every deeper descendant. This single field is the main mechanism by which the taxonomy database drives the rest of NCBI, and it accepts any name at any rank, plus synonyms that map to the same taxid.

## How NCBI Builds and Maintains the Tree

Understanding the curation model prevents the most common misreading of the database. NCBI Taxonomy is not assembled by a committee voting on one grand tree. It is a synthesis that draws on the taxonomic literature, on the judgment of individual taxonomists who serve as advisers, and on partner resources that maintain specialist classifications. When a GenBank submission deposits a sequence under a new species name, the taxonomy is extended to accommodate it, usually by anchoring the new name to an existing genus or family. When published work revises a group, the corresponding nodes are updated.

Three practical consequences follow from this model. First, the tree tracks published consensus with a lag, so a very recent taxonomic rearrangement may not appear immediately. Second, groups with active molecular revision, such as fungi and many invertebrate lineages, change more often than groups with stable taxonomy. Third, names deposited in GenBank under old classification schemes persist in the sequence records even after the taxonomy node is updated or merged, which is why synonym mapping matters in any large-scale retrieval.

## A Practical NCBI Databases Guide: Six Steps From Taxonomy to Sequences

![Six numbered steps running from locating an organism in the Taxonomy Browser to interpreting a tree against the taxonomy.](/article-images/illustrations/ncbi-taxonomy-evolutionary-relationships-six-step-workflow.webp)

This section is the hands-on core of using ncbi taxonomy to explore evolutionary relationships. Each step builds on the last, and the command-line examples use the **Entrez Direct** utilities (the `esearch` and `efetch` programs installed with the E-utilities package), which are the official programmatic interface to NCBI.

### Step 1. Locate your organism and read its lineage

Start in the Taxonomy Browser. Search for your organism, confirm you have the right one (synonyms and similarly spelled names collide more often than you expect), and read the lineage top to bottom. Note the taxid, the parent genus, and any unranked nodes between your species and the next formal rank. If the browser lists multiple names, check which one is the currently accepted scientific name and which are synonyms.

On the command line, the same lookup looks like this:

```
esearch -db taxonomy -term "Homo sapiens" | efetch -db taxonomy -format xml
```

This returns the full record, including taxid, lineage, and rank information, in a form your scripts can parse.

### Step 2. Expand to the group you want to compare

Rarely do you care about one species. You care about a genus, a family, or all sequences from a host group. Two tools handle this.

The **Common Tree** tool in the Taxonomy Browser takes a list of taxids or names and returns the smallest subtree containing all of them, with the branching structure NCBI records among them. Paste a species list from your study, get the tree, and you immediately see which of your taxa are redundant (two names mapping to the same taxid) and which fall in surprising places. You can save the result as a token tree for [downstream processing](/knowledge/molecular-biology/downstream-processing).

For scripted retrieval, combine the `[Organism]` filter with Entrez:

```
esearch -db nucleotide -term "Eriobotrya[Organism] AND internal transcribed spacer[All Fields]" | efetch -db nucleotide -rettype fasta -retmode text
```

Because the filter expands down the whole subtree, you get every descendant species automatically. This is the standard pattern for harvesting homologous sequences across a clade.

### Step 3. Survey what data actually exist

Before planning a phylogeny, check data availability per taxon. The Taxonomy Browser shows record counts per node. The NCBI Datasets interface (ncbi.nlm.nih.gov/datasets) provides a cleaner programmatic view of genome and taxonomy records, for example:

```
datasets summary taxonomy taxon "Rhaphiolepis"
```

A quick count of genome assemblies, mitochondrial genomes, or transcriptomes per species tells you whether a genome-scale analysis is realistic or whether you will be working with a few loci. Many published phylogenies are constrained exactly this way. The Truncatelloidea mitogenomic study discussed below could integrate 20 genera across 11 families only because enough mitochondrial genomes already existed in NCBI, and it still had to add 15 newly sequenced mitochondrial genomes to fill the gaps [1].

### Step 4. Retrieve the sequences

With the scope defined, pull sequences through Entrez, using the taxonomy filter to stay inside your clade. For genome-scale projects, the SRA (Sequence Read Archive) is the source of raw reads. The sesame genomics study discussed below combined five newly resequenced Egyptian genotypes with 19 publicly available sesame datasets from NCBI SRA to build a 24-genotype comparison, which is the standard pattern for turning one lab's project into a population-scale analysis [2].

### Step 5. Annotate, align, and build the tree

The taxonomy database plays a supporting role here but a vital one. Taxids and consistent species names are what let you color the resulting tree by taxonomy, collapse nodes by family, and detect mislabeled sequences. During quality control, a sequence that clusters far from its assigned taxonomy is either an interesting find or a misidentification, and both outcomes matter. Public-data mining projects lean on this check constantly. The Myochromella survey discussed below screened publicly available ITS sequences and amplicon data at a 98 percent similarity threshold and found that a large fraction of the recorded diversity had no formal species designation at all [3].

### Step 6. Interpret the tree against the taxonomy

The final step is interpretation, which deserves its own section below because it is where most errors happen.

## A Worked Example: From Taxonomy Nodes to a Published Phylogeny

Consider a question that mirrors real published work. Loquat (*Eriobotrya*) and *Rhaphiolepis* are two closely related genera in the rose family tribe Maleae, and the boundary between them has been contested. A 2025 study reassessed their phylogeny by combining nuclear ribosomal DNA, [mitochondrial DNA](/blog/guides/mitochondrial-dna), and micromorphological characters across 53 accessions, drawing on 16 nrDNA sequences and one mitochondrial sequence acquired from NCBI [4].

Walk through how NCBI Taxonomy supports a project like this:

1. **Frame the sampling.** The Taxonomy Browser confirms both genera sit within Maleae, Rosaceae, which tells you the relevant outgroups (other Maleae genera) and where the root of your ingroup comparison should go.
2. **Retrieve homologs.** An `[Organism]`-filtered Entrez query on each genus plus related Maleae genera pulls available nrDNA and mitochondrial sequences, with taxids attached to every record.
3. **Fill gaps.** Where NCBI sequences are missing or unreliable, the authors sequenced new accessions, exactly as the study did for most of its 53 accessions [4].
4. **Analyze.** Alignments of nrDNA and concatenated mitochondrial protein-coding genes feed into maximum likelihood and Bayesian analyses, with morphology as an independent line of evidence.
5. **Interpret against the taxonomy.** The nrDNA tree and leaf micromorphology supported *Rhaphiolepis* as monophyletic with high support and *Eriobotrya* as monophyletic with moderate to low support, and additional characters from petioles, fruits, and fruit apical sepals supported treating the two as separate genera [4].

Notice the division of labor. NCBI Taxonomy defined the frame, supplied some sequences, and provided the naming framework the conclusions were expressed in. The evolutionary claims themselves came from the authors' own phylogenetic analysis. That division is the essence of using ncbi taxonomy to explore evolutionary relationships correctly.

## Interpreting Evolutionary Signals: Clades, Sister Groups, and Conflicts

### The vocabulary you need

A **clade** is an ancestor and all of its descendants. A **sister group** relationship means two taxa share an immediate common ancestor not shared with any other taxon in the analysis. **Monophyly** means a named group contains exactly one complete clade. When a published tree "strongly supports the monophyly" of families, as the Truncatelloidea study reported for most of the families and genera it analyzed [1], it means the molecular data recovered each family as a single complete clade, consistent with their placement in NCBI Taxonomy. When a study reports a family as polyphyletic or a genus as paraphyletic, it means the data disagree with the current classification, and taxonomic revision often follows.

### Gene trees versus species trees

The single most consequential idea in interpretation is that a tree built from one gene or one region is a **gene tree**, and it need not match the history of the species. Several forces drive the difference. **Incomplete lineage sorting** allows alleles to persist through speciation events so a gene's history branches differently from the species history. **[Gene duplication](/knowledge/molecular-biology/gene-duplication)** creates paralogous copies whose relationships reflect duplication events, not organismal splits. **Introgression** moves genes across species boundaries.

The IGKC study makes this concrete. Immunoglobulin genes are an ancient family with repeated duplications, and the authors expanded sampling from 16 fully characterized mammal species in the specialist IMGT database to 124 IGKC sequences across 104 mammals, using genomic sequences retrieved from NCBI and Ensembl [5]. They found independent duplications of IGKC in two lagomorph lineages, generating rabbit IGKC1 and IGKC2 [5]. A naive species-level reading of a single IGKC tree would be wrong for exactly those lineages, because the gene tree contains duplication events the species tree does not. The same study reported pervasive purifying selection across the gene with individual codons under positive selection and episodic diversifying selection in several lineages [5], which is another layer of signal that only appears once sequences are correctly homologized within the taxonomy framework.

### Conflicting markers are information

The Eriobotrya study also illustrates why multiple markers matter. Its mitochondrial gene tree, built from 52 protein-coding genes, offered limited evolutionary resolution because of low sequence variability and displayed incongruence with the nuclear ribosomal DNA results [4]. A beginner reads that as failure. An experienced analyst reads it as a finding. Slow-evolving markers cannot resolve recent divergences, and organelle trees can disagree with nuclear trees for legitimate biological reasons. NCBI Taxonomy helps you detect this early, because marker availability per taxon is visible before you commit to a dataset.

### What NCBI Taxonomy itself does and does not encode

![Side by side comparison of the NCBI taxonomy tree and a computed phylogeny across branch lengths, support values and unresolved nodes.](/article-images/illustrations/ncbi-taxonomy-evolutionary-relationships-tree-versus-phylogeny.webp)

The taxonomy tree carries no branch lengths, no bootstrap values, and no node support figures. Those exist only in the underlying phylogenetic analyses. When the browser draws a clean binary fork between two groups, it reflects curation of published evidence, not a computed confidence value. Treat every internal node in the taxonomy as "accepted current view," and reserve statements about strength of support for your own analyses.

## NCBI Taxonomy Across the Tree of Life: Research Examples

The same workflow flexes across vastly different organisms, scales, and questions. These examples, all drawn from studies that built their analyses on NCBI data, show the range.

**Virus surveillance and lineage dynamics.** Zika virus phylogenetics depends entirely on community sequence deposition. One 2026 analysis examined 1,889 ZIKV genomic deposits in NCBI, found that the majority originated from the Americas (59 percent) and Asia (27 percent), with the Asian lineage predominating and the West African lineage most prevalent in Africa, and showed that field-collected mosquito eggs carried the East African lineage, suggesting mosquito eggs may serve as viral reservoirs [6]. By combining new C-prM fragment sequencing from clinical and mosquito samples with NCBI sequences, the authors placed their samples in a phylogenetic tree and detected clustering with African lineages, indicating a potential introduction of those lineages outside Africa [6]. The taxonomic and lineage framework embedded in NCBI is what makes a global picture like this computable from thousands of independent submissions.

**Genotype structure in a pathogen.** A study of tick-borne encephalitis virus retrieved 263 genome sequences from the NCBI database and classified them into five genotypes (Baikalian, European, Far-Eastern, Himalaya, and Siberian), with whole-ORF nucleotide similarity ranging from 81.5 to 88.0 percent and amino acid similarity from 93.0 to 96.4 percent across genotypes [7]. Within that framework, entropy analysis located highly variable sites per genotype, positive selection analysis identified sites under selection, and the authors consolidated variable sites, shared genotype-specific mutations, and selected sites into 37 key amino acid positions, mostly on protein surfaces, with potential functional and pathogenicity relevance requiring further validation [7]. Here the taxonomy-level organization into lineages is the scaffold on which mutation-level evolutionary analysis hangs.

**Population and cultivar phylogenetics in crops.** Whole-genome resequencing of five Egyptian sesame genotypes, integrated with 19 public sesame datasets from NCBI SRA, yielded 6,106,085 nucleotide variants across 24 genotypes and, after filtering for coverage, 26,424 SNPs with minor allele frequency above 0.05 for downstream analysis [2]. Population structure analysis identified three genetic clusters, with Egyptian cultivars forming a genetically pure separate cluster, and phylogenetic analysis supported country-based clustering [2]. Notice that "phylogenetic analysis" at this scale is a neighbor-joining or tree-building step over SNP data, sitting one rung below species-level phylogenetics on the same conceptual ladder, and still anchored to the taxonomy of a single species.

**Gene family evolution across a class.** A survey of avian cathelicidin immune genes mined genome assemblies in the NCBI database for 72 species across 26 orders, mapping cluster composition, gene structure, and phylogenetic relationships [8]. The authors identified 190 putative cathelicidin genes, of which 103 were previously undescribed, and showed lineage-specific patterns such as Galliformes carrying cath1, a putative pseudogene of cathB1 in Anseriformes, and cath3 lacking or pseudogenized in Passeriformes and Falconiformes [8]. This is taxonomy-driven [comparative genomics](/blog/guides/comparative-genomics) in its purest form. The species list came straight from the NCBI tree, and every evolutionary claim (duplications, losses, pseudogenization) was made against that framework.

**Virus discovery in a crop species.** A virome study of cacao analyzed 109 publicly available RNA-seq libraries from nine BioProjects using a pipeline of viral sequence enrichment, hybrid assembly with multiple assemblers, and similarity searches against NCBI non-redundant databases, identifying ten putative novel viruses associated with the cocoa microbiome and a novel Badnavirus species [9]. Taxonomy enters twice here. First, the host taxonomy fixes the sampling frame (Theobroma cacao and its associated BioProjects). Second, the viral hits are classified by similarity against curated reference databases, and classification decisions feed back into the taxonomy tree as new viral sequences are deposited.

**Fungal diversity from environmental amplicons.** A study of the mushroom genus Myochromella combined ITS sequences from NCBI-GenBank with fungal amplicon data from NCBI-SRA, applying a 98 percent sequence similarity threshold to identify eleven species, eight of them without formal designations, and found the genus primarily Northern Hemisphere-centric with a tendency toward endemic distribution [3]. This workflow, mining unstructured amplicon reads under a defined similarity threshold and mapping them onto taxonomy nodes, is now one of the most common ways cryptic diversity is documented.

Four lessons generalize from these examples. Taxonomy defines the sampling frame. Public deposits turn single-lab projects into clade-scale comparisons. Lineage frameworks in the database make global patterns computable. And interpretation always separates the taxonomy scaffold from the evolutionary inference built on it.

## How to Use NCBI Effectively in Your Own Workflow

A few habits separate efficient analysts from frustrated ones.

**Start every project at the Taxonomy Browser, not the sequence databases.** Confirm the accepted name, the taxid, and the position of your organisms in the tree before writing a single query. An hour here prevents days of rework caused by wrong-name queries.

**Use the subtree expansion deliberately.** `[Organism]` filters expand to all descendants, which is powerful when you want a clade and dangerous when you think you filtered to one species but actually captured its whole genus. Check the returned taxids.

**Script against taxids, validate against names.** Taxids are the stable machine keys, but they change over time. Any long-running pipeline should check merged.dmp and delnodes.dmp periodically, or re-resolve names to taxids at each run.

**Respect rate limits in E-utilities.** The official guidance asks for no more than three requests per second without an API key, and scripts should post their email address in the tool parameter. These are documented in the NCBI E-utilities documentation and exist to keep the service available to everyone.

**Pair every taxonomic claim with an analytical claim.** Saying two taxa are "related" because they share a genus is a taxonomy statement. Saying they are sister lineages with high bootstrap support in your nrDNA alignment is an evolutionary statement, and the published literature supports distinguishing the two [4].

## Common Mistakes and Limitations

**Treating the taxonomy tree as a phylogram.** The displayed tree has no branch lengths and no support values. Reading deep or shallow branches off the browser graphic is meaningless. Branch lengths come from your own or others' explicit analyses.

**Assuming equal rank means equal age.** A genus of beetles and a genus of birds can differ in age by tens of millions of years. Rank labels are classification conventions, not time stamps. Cross-clade comparisons based on rank alone are unreliable, and this misconception quietly corrupts many comparative datasets.

**Hard-coding taxids without change management.** Taxids get merged and deleted. Pipelines that do not consult merged.dmp and delnodes.dmp silently lose or misattribute samples. Re-resolve names to current taxids on a schedule.

**Ignoring sampling bias.** Record counts per node reflect decades of uneven research effort. Model organisms and disease vectors are saturated, while many tropical lineages have a handful of records. Any "diversity" pattern computed directly from NCBI counts conflates biology with sampling effort, a point the Myochromella work navigated by defining explicit similarity thresholds before counting species [3].

**Confusing gene trees with species trees.** As the IGKC duplications in lagomorphs show, gene family history within a well-defined mammalian taxonomy can contain duplication events that make single-gene trees misleading for species relationships [5]. Orthology checks are not optional for gene family studies.

**Expecting every node to be resolved.** Polytomies in the taxonomy are honest representations of unresolved relationships. Forcing a binary reading, or assuming the browser order within a polytomy is meaningful, manufactures signal that does not exist.

**Assuming sequence labels are correct.** Misidentified specimens and mislabeled sequences exist in GenBank. A sequence whose placement contradicts its taxonomy label is a flag to investigate, not a fact to discard blindly or accept blindly.

**Forgetting that taxonomy lags the literature.** A classification published last month may not be reflected yet. For groups under active revision, check the primary literature alongside the database, the way the Rosaceae study integrated fresh nrDNA data with existing NCBI records to test a contested genus boundary [4].

## Quick Review

- NCBI Taxonomy assigns every organism in GenBank a unique taxid, a rank, and a lineage, and it indexes all sequence and genome data by that identifier.
- The taxonomy tree is a curated consensus classification, not a phylogram. It has no branch lengths and no support values, and unresolved nodes appear as polytomies.
- The `[Organism]` Entrez filter expands down the whole subtree, which is the main bridge between the taxonomy database and every other NCBI database.
- Taxdump files (names.dmp, nodes.dmp, merged.dmp, delnodes.dmp) let you script the entire tree, and merged and deleted taxids must be handled in any pipeline.
- Gene trees can disagree with species trees through duplication, incomplete lineage sorting, and introgression, so gene family studies need orthology checks before evolutionary interpretation.
- Marker choice matters. Slow-evolving markers may fail to resolve recent divergences and can conflict with nuclear data, as shown in mitochondrial versus nrDNA comparisons.
- Record counts per taxon reflect sampling effort, not diversity. Establish explicit thresholds and quality criteria before counting anything.

## Frequently Asked Questions

### What does NCBI taxonomy evolutionary relationships mean?

It means using the curated NCBI classification tree, which links every organism with sequence data to a taxid and a lineage, as the framework for studying relatedness among organisms. The tree captures current scientific consensus about how organisms are grouped, and you test and refine specific evolutionary claims by retrieving sequences through that framework and running your own phylogenetic analyses.

### Is the NCBI taxonomy tree a real phylogeny?

It is a consensus classification informed by phylogenetic evidence, not a computed phylogeny of every node. The tree contains no branch lengths or support values, and it includes polytomies where relationships are unresolved. Treat it as an authoritative naming and grouping framework, and rely on explicit phylogenetic analyses for branch lengths and confidence statements.

### How do I find the taxonomy ID for my organism?

Search the organism name at the NCBI Taxonomy Browser, and the resulting record shows the taxid, the accepted scientific name, synonyms, and the full lineage. Programmatically, use the E-utilities, for example `esearch -db taxonomy -term "Homo sapiens"` piped into `efetch -db taxonomy -format xml`, which returns the taxid and lineage in machine-readable form.

### How do I download a subtree of the taxonomy tree?

Use the Common Tree tool in the Taxonomy Browser to enter a list of names or taxids and export the minimal subtree containing them. For complete programmatic access, download the taxdump archive from the NCBI FTP site and parse nodes.dmp and names.dmp, which together let you reconstruct any subtree, including all parent-child relationships and ranks.

### Can I use NCBI Taxonomy to build my own phylogenetic tree?

Yes, as the backbone of the project. Use the taxonomy to define your sampling frame, retrieve homologous sequences through taxonomy-filtered Entrez queries, verify sequence labels against their taxids during quality control, and then build the tree yourself with alignment and phylogenetic software. The taxonomy provides the frame and the data index, while the tree-building and support values come from your analysis.

### Why did a species name or taxid in NCBI Taxonomy change?

Names and taxids change when taxonomic research revises a group, synonymizing one name into another, splitting one species into several, or reclassifying a genus within a different family. NCBI records these events in the merged.dmp and delnodes.dmp files, so old taxids either redirect to their replacements or are marked deleted. Check those files in any long-running pipeline and re-resolve names before each analysis run.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What does NCBI taxonomy evolutionary relationships mean?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "It means using the curated NCBI classification tree, which links every organism with sequence data to a taxid and a lineage, as the framework for studying relatedness among organisms. The tree captures current scientific consensus about how organisms are grouped, and you test and refine specific evolutionary claims by retrieving sequences through that framework and running your own phylogenetic analyses."
      }
    },
    {
      "@type": "Question",
      "name": "Is the NCBI taxonomy tree a real phylogeny?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "It is a consensus classification informed by phylogenetic evidence, not a computed phylogeny of every node. The tree contains no branch lengths or support values, and it includes polytomies where relationships are unresolved. Treat it as an authoritative naming and grouping framework, and rely on explicit phylogenetic analyses for branch lengths and confidence statements."
      }
    },
    {
      "@type": "Question",
      "name": "How do I find the taxonomy ID for my organism?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Search the organism name at the NCBI Taxonomy Browser, and the resulting record shows the taxid, the accepted scientific name, synonyms, and the full lineage. Programmatically, use the E-utilities, for example esearch -db taxonomy -term \"Homo sapiens\" piped into efetch -db taxonomy -format xml, which returns the taxid and lineage in machine-readable form."
      }
    },
    {
      "@type": "Question",
      "name": "How do I download a subtree of the taxonomy tree?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Use the Common Tree tool in the Taxonomy Browser to enter a list of names or taxids and export the minimal subtree containing them. For complete programmatic access, download the taxdump archive from the NCBI FTP site and parse nodes.dmp and names.dmp, which together let you reconstruct any subtree, including all parent-child relationships and ranks."
      }
    },
    {
      "@type": "Question",
      "name": "Can I use NCBI Taxonomy to build my own phylogenetic tree?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Yes, as the backbone of the project. Use the taxonomy to define your sampling frame, retrieve homologous sequences through taxonomy-filtered Entrez queries, verify sequence labels against their taxids during quality control, and then build the tree yourself with alignment and phylogenetic software. The taxonomy provides the frame and the data index, while the tree-building and support values come from your analysis."
      }
    },
    {
      "@type": "Question",
      "name": "Why did a species name or taxid in NCBI Taxonomy change?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Names and taxids change when taxonomic research revises a group, synonymizing one name into another, splitting one species into several, or reclassifying a genus within a different family. NCBI records these events in the merged.dmp and delnodes.dmp files, so old taxids either redirect to their replacements or are marked deleted. Check those files in any long-running pipeline and re-resolve names before each analysis run."
      }
    }
  ]
}
</script>

## Sources

1. [Mitogenomic phylogeny of Truncatelloidea with description of Aenigmula sinensis Tang, Han & Kong, sp. nov. (Mollusca, Gastropoda, Littorinimorpha, Truncatelloidea).](https://pubmed.ncbi.nlm.nih.gov/42211056/)
2. [Exploring the genetic landscape of sesame: whole genome sequencing reveals distinct Egyptian lineages and SNP variability.](https://pubmed.ncbi.nlm.nih.gov/41619030/)
3. [Myochromella unveiled: exploring its global distribution through a public database of amplicons.](https://pubmed.ncbi.nlm.nih.gov/40413396/)
4. [Reassessing the Evolutionary Relationships of Eriobotrya and Rhaphiolepis (Rosaceae): Evidence from Micromorphology, Complete Nuclear Ribosomal DNA and Mitochondrial Genomic Data.](https://pubmed.ncbi.nlm.nih.gov/41463513/)
5. [The evolutionary history of IGKC in mammals reveals ancient duplications and remarkable divergence in lagomorphs.](https://pubmed.ncbi.nlm.nih.gov/41169362/)
6. [ZIKV Phylogenetic Characterization Reveals Evolutionary Diversity, Regional Dissemination, and Emergence of African Lineages in Brazil.](https://pubmed.ncbi.nlm.nih.gov/41157841/)
7. [Evolutionary Patterns and Genotype-Specific Amino Acid Mutations of Tick-Borne Encephalitis Virus.](https://pubmed.ncbi.nlm.nih.gov/39940723/)
8. [Exploring diversity in avian immune defence: Insights from cathelicidin clusters.](https://pubmed.ncbi.nlm.nih.gov/40127729/)
9. [Theobroma cacao Virome: Exploring Public RNA-Seq Data for Viral Discovery and Surveillance.](https://pubmed.ncbi.nlm.nih.gov/40431635/)