# NCBI Taxonomy vs. Other Taxonomic Databases: Which One Should You Use NCBI Taxonomy is a curated classification backbone that assigns a stable identifier, the taxid, to every organism named in GenBank, RefSeq, and the Sequence Read Archive, while databases such as GTDB, BOLD, and MGnify rebuild classification around a single data type, namely whole genomes, COI barcodes, or metagenomic catalogs. The rule of thumb is simple. Use NCBI Taxonomy as the default backbone whenever your workflow touches NCBI sequence data or tools that read taxids, and bring in a specialist database when your question needs something the backbone does not supply, such as phylogeny-consistent prokaryotic names from GTDB or the barcoding species baseline from BOLD. The choice is not cosmetic. In a dairy cattle metagenomics study, the same shotgun reads were classified against MGnify, GTDB, and NCBI, and each database produced its own genus and species feature sets for downstream analysis [1]. Names are the join keys of a bioinformatics pipeline. When two databases disagree about a join key, every result downstream of the join changes. This guide to NCBI databases and their alternatives explains what NCBI Taxonomy actually is, puts it side by side with the main specialists, shows how to use NCBI in concrete workflows with real commands, and closes with worked scenarios, the mistakes that cost the most time, and direct answers to the questions that come up most often. ## What NCBI Taxonomy Is (and What It Is Not) A taxonomy database stores organism names, their synonyms, and a parent-child hierarchy that places every name from the root of life down to strain-level nodes. NCBI Taxonomy performs that job for the entire NCBI sequence ecosystem. GenBank, the public archive of submitted sequences, RefSeq, the curated non-redundant reference set, and SRA, the Sequence Read Archive that stores raw reads, all attach a taxid to every record. The non-redundant (nr) BLAST database, which merges records from these sources into a single search target, inherits the same taxonomy, which is why a similarity search against nr returns an organism name for every hit [2]. Every node in NCBI Taxonomy carries a taxid, a stable integer identifier. Homo sapiens is taxid 9606, Escherichia coli is taxid 562, and the remaining nodes cover everything from uncultured archaeal lineages to individual laboratory strains and virus isolates. NCBI taxonomists curate the database directly. They merge duplicate names, track synonymy, flag uncertain placements, and update the classification as formal nomenclature changes. When two taxids are merged, the old identifier is preserved in a merge table so that scripts written against the old number can be repaired rather than silently broken. What NCBI Taxonomy is not matters as much as what it is. It is not a computed phylogeny. It is a consensus classification designed to organize sequence records, and its internal arrangement does not always match the best current phylogenetic hypothesis for a group. If you need a tree, build one from aligned sequences. It is not a quality filter for organism labels. GenBank records carry the organism name the submitter provided, and mislabeled sequences in public archives such as GenBank are a documented, recurring problem [3]. It is not genome-only. The taxonomy spans viruses, bacteria, archaea, fungi, plants, and animals, including the BioProjects (NCBI records that group related datasets under one project) and raw-read libraries that virome studies pull from SRA [2]. That breadth is exactly why so many cross-domain pipelines depend on it. ## The Main Alternatives, on Their Own Terms ### GTDB: Taxonomy Rebuilt from Genomes The Genome Taxonomy Database (GTDB, at gtdb.ecogenomic.org) takes whole bacterial and archaeal genomes, infers a phylogeny from a standardized set of conserved marker genes, and then defines the taxonomy so that every name sits on that tree. Where no formal name exists for a lineage, GTDB assigns a standardized placeholder rank, so every genome receives a complete classification from domain to species. The companion tool GTDB-Tk places a new genome into that classification automatically. The consequence is that GTDB names and NCBI names do not always agree, because the two databases answer different questions. NCBI Taxonomy records the names in use across the sequence archives. GTDB asks what a consistent genome phylogeny supports and accepts that some of its names are placeholders awaiting formal description. This is why the dairy metagenomics study treated MGnify, GTDB, and NCBI as three separate classification systems rather than three views of one truth [1]. GTDB covers prokaryotes only, and only those with genomes. If your study includes eukaryotes or viruses, GTDB cannot be your backbone. ### BOLD and BINs: The Barcoding Baseline The Barcode of Life Data System (BOLD) centers on cytochrome c oxidase subunit I (COI), the standard barcode gene for animals. Its signature contribution is the Barcode Index Number (BIN), an algorithmic cluster of COI sequences that acts as a proxy for a species. BINs give barcoding and environmental DNA (eDNA, DNA sampled directly from water, soil, or other substrates rather than from an organism) studies a standardized taxonomic baseline that does not depend on the organism labels attached to sequences in GenBank. Coverage gaps in both systems are real. GenBank and BOLD are the two main repositories for COI barcodes, and until very recently the nightjar louse fly Pseudolynchia garzettae had no sequences in either one, while every sequenced specimen of its genus matched the single species Pseudolynchia canariensis. A newly generated 662 bp COX1 sequence was what confirmed a Welsh specimen as a separate species [4]. If your organism is obscure, plan to generate your own barcode rather than assuming one exists. ### MGnify: Metagenomic Catalogs from EMBL-EBI MGnify is the metagenomics resource of EMBL-EBI, the European archive organization that parallels NCBI. Its distinguishing feature is the catalog, a standardized, periodically rebuilt collection of genomes and genes assembled from metagenomes of a defined environment. The cow rumen catalog used in the dairy study is one example [1]. Classifying reads against a catalog gives you genus and species feature sets that are internally consistent, because every genome in the catalog was assembled, dereplicated, and annotated the same way. The trade-off is scope. A catalog covers what the catalog contains, and its genus and species definitions belong to the catalog, not to NCBI. Results against MGnify catalogs and results against NCBI RefSeq are both defensible, but they are not interchangeable. ### Curated rRNA Databases for Amplicons For amplicon work (PCR-amplified marker regions such as the 16S [ribosomal RNA](/knowledge/bioinformatics/ribosomal-rna-structure-taxonomic-profiling) gene or the internal transcribed spacer, ITS), the community leans on curated, marker-specific reference databases such as SILVA for ribosomal RNA, UNITE for fungal ITS, and PR2 for protists. These trade breadth for quality through manual curation and consistent alignments, and they are the standard references for assigning amplicon sequence variants (ASVs, exact sequence variants) or operational taxonomic units (OTUs, sequences clustered at a similarity threshold). Amplicon studies built directly on NCBI sources follow the same logic of explicit thresholds. One macrofungus study mined ITS sequences from NCBI-GenBank together with fungal amplicon data from NCBI-SRA and clustered at a 98% similarity threshold, which resolved eleven species in the genus Myochromella, eight of them lacking formal names [5]. The threshold was doing taxonomic work, and the authors said so. ### When the NCBI and EBI Pair Beats Either Alone NCBI and EBI resources overlap heavily, and many projects use both. A sepsis meta-analysis merged microarray samples from NCBI [Gene Expression Omnibus](/blog/guides/gene-expression-omnibus) and EMBL-EBI ArrayExpress into a single meta-expression set, because neither archive alone held enough matched whole-blood samples for classifier training [6]. The same complementarity applies to taxonomy-adjacent work. NCBI gives you the taxid backbone and the archives, and EBI resources such as MGnify give you environment-specific catalogs built with uniform processing. ## NCBI Taxonomy vs. Other Taxonomic Databases, Side by Side The difference between NCBI taxonomy and other databases comes down to what each one is built from and what each one promises. The table below is the core of this comparison. | Criterion | NCBI Taxonomy | GTDB | BOLD | MGnify | SILVA, UNITE, PR2 | |--|--|--|--|--|--| | Built from | Organism records across GenBank, RefSeq, and SRA | Bacterial and archaeal genomes | COI barcode sequences | Metagenomic assemblies and catalogs | Curated rRNA and ITS alignments | | Classification basis | Curated consensus names maintained by NCBI taxonomists | Phylogeny of conserved marker genes, taxonomy aligned to the tree | Clustering of COI sequences into BINs | Catalog membership under one pipeline | Manual curation with consistent alignments | | Coverage | All domains of life plus viruses | Bacteria and archaea with genomes | Eukaryote barcode groups, strongest for animals | Mainly prokaryotes from metagenomes | Organisms represented in rRNA marker space | | Standout strength | Universal taxids that BLAST, Kraken2, Centrifuge, and NCBI tools read natively | Complete rank-by-rank names even for unnamed lineages | A species baseline independent of submitter labels | Internally consistent genus and species features | High-quality marker references for amplicons | | Main watch-out | Submitter labels can be wrong [3], and RefSeq genomes can carry plasmids that skew species calls [7] | Names differ from formal nomenclature and from NCBI names [1] | COI only, and many species have no records yet [4] | Scope limited to the catalog | Marker specific, not genome aware | | Best for | Default backbone, reporting, cross-domain pipelines | Prokaryotic genome classification | eDNA and barcoding studies | Environment-specific metagenomic features | 16S, 18S, and ITS amplicon assignment | Tool compatibility decides many projects before any biology does. | Tool or resource | Taxonomy it reads natively | |--|--| | BLAST with taxid mapping | NCBI taxids | | Kraken2, Centrifuge | NCBI taxids | | NCBI datasets and E-utilities | NCBI taxids | | GTDB-Tk | GTDB | | BOLD API | BINs | | MGnify analyses | MGnify catalogs | Three criteria decide most real choices between NCBI taxonomy or other databases. Naming philosophy. NCBI records the names in use, GTDB derives names from a genome tree, BOLD clusters barcodes into species proxies, and MGnify defines taxa by catalog membership. None of these is wrong. They answer different questions, and mixing them without a mapping step is where projects go wrong. Coverage versus depth. NCBI Taxonomy is the broadest, spanning all domains and viruses. The specialists go deeper inside their scope. GTDB gives every prokaryotic genome a complete rank-by-rank classification, BOLD gives animal barcodes an algorithmic species baseline, and MGnify gives metagenomes a consistent catalog. Choose breadth when your pipeline is heterogeneous and depth when your question is narrow. Tool compatibility. BLAST, Kraken2, Centrifuge, the datasets tool, and the E-utilities all read NCBI taxids natively. GTDB-Tk reads GTDB. BINs live inside BOLD and its API. If your downstream tooling expects taxids, starting from another database means converting everything, and conversion is where silent errors creep in. ## How to Use NCBI Taxonomy in Real Workflows ### Look Up Names and Taxids The Taxonomy Browser at https://www.ncbi.nlm.nih.gov/taxonomy is the fastest way to check a name, see its rank, and read its taxid. For scripted work, the E-utilities (installed as the EDirect command line tools) query the same database: ``` esearch -db taxonomy -query "Escherichia coli[sciname]" | efetch -db taxonomy -format xml ``` That pipeline returns the full record for taxid 562, including the lineage and the synonyms NCBI maintains for the name. For bulk work, download the new_taxdump archive from https://ftp.ncbi.nlm.nih.gov/pub/taxonomy/new_taxdump/. Two files do most of the lifting. nodes.dmp holds the parent-child structure, and names.dmp holds the names and name classes. Joining a results table on taxid against these two files gives you lineage-aware summaries without any web calls. ### Attach Taxonomy to BLAST Databases BLAST databases can carry taxonomy directly. Build one with makeblastdb, passing a two-column mapping file that lists each sequence accession and its taxid: ``` makeblastdb -in refseq_bacteria.fna -dbtype nucl -parse_seqids -taxid_map taxmap.txt -blastdb_version 5 blastn -query reads.fna -db refseq_bacteria -outfmt "6 qseqid staxids pident length" -max_target_seqs 10 ``` The staxids field returns the taxid of each subject hit, so you can summarize results by organism without a separate lookup step. If you want scientific names inside the output itself, install the taxdb files that accompany NCBI BLAST databases and add the sscinames field to the output format. The same setup doubles as a label-audit tool. When a GenBank label looks suspicious, re-BLAST the sequence against a broad database and read the taxids of the top hits. A sequence labeled as one species that clusters entirely inside another taxon is a mislabel candidate, and mislabels are common enough that auditing is standard practice [3]. ### Classify Metagenomic Reads ![Flowchart from building a classifier database through read classification to a plasmid check and either a confirmed call or a reassigned taxon.](/article-images/illustrations/ncbi-taxonomy-vs-other-databases-metagenomics-branch.webp) k-mer-based classifiers such as Kraken2 and Centrifuge are built directly on NCBI Taxonomy. They index a reference set, typically RefSeq genomes, into k-mers (short subsequences of fixed length), and assign each read the lowest common ancestor (LCA, the most specific taxon shared by all matches) of its matching k-mers. That design is why a metagenomic report arrives as a tree of NCBI taxids. That coupling is also where a documented failure mode lives. In a comparison of aerobic culture, PCR, and shotgun metagenomics for Salmonella enterica in cattle feces, k-mer classification against RefSeq flagged 11 of 60 samples as containing S. enterica reads, while culture and PCR agreed with each other on 3 of 60. The excess traced to plasmid sequences (extrachromosomal DNA elements) stored inside bacterial RefSeq genomes, which uniquely associated shared plasmid DNA with S. enterica. After separating plasmid sequences from the genomes and reassigning them to their own taxon, the metagenomic call dropped to 6 of 60 [7]. The lesson is not that the classifier failed. The lesson is that taxonomy attached to genomes inherits every quirk of how those genomes were submitted. ### Retrieve Sequences for Primer and Probe Design Assay design starts at the taxonomy node. An H5 and H7 influenza detection assay pulled conserved hemagglutinin sequences from the NCBI database, aligned them to find regions conserved within the target subtypes, designed primers, probes, and CRISPR RNAs (crRNAs, the guide sequences used by Cas12a detection), and then checked candidate primers with Primer-BLAST, NCBI's tool for testing primers against a sequence database [8]. The generic recipe is the same for any pathogen or gene family. Download sequences by taxid, align, identify regions conserved within the target group and absent from near neighbors, and validate specificity with Primer-BLAST at https://www.ncbi.nlm.nih.gov/tools/primer-blast/. ### Pull Genomes for Phylogenetics and Comparative Genomics Comparative studies routinely combine newly generated genomes with everything NCBI already holds. A [Trueperella pyogenes](/knowledge/bacteria/livestock-bacteria/trueperella-pyogenes) study sequenced 37 Danish isolates de novo and compared them against all 126 genomes of sufficient quality available in NCBI using core-genome SNP analysis, which counts single nucleotide polymorphisms (SNPs) across the set of genes shared by all strains. The NCBI-wide comparison revealed two main clusters that followed host species, with pig strains and cattle strains on separate branches [9]. The same pattern holds in fungi and mollusks. A Boletaceae study combined eight newly sequenced mitochondrial genomes (mitogenomes) with 28 additional mitogenomes from NCBI and recovered strongly supported phylogenies [10]. A Truncatelloidea study integrated 15 new mitogenomes with existing NCBI data spanning 20 genera in 11 families [11]. The datasets command line tool pulls a taxon's genomes in one call: ``` datasets download genome taxon "Trueperella pyogenes" ``` Filter by assembly quality before analysis, exactly as the Trueperella study filtered to genomes of sufficient quality [9]. A genome with thousands of contigs (fragmentary assembly pieces) will scatter spurious SNPs across a core-genome alignment. ### Mine Assemblies, SRA, and Annotations Genome mining starts from the taxonomy tree. An avian immunology study walked genome assemblies for 72 bird species across 26 orders in NCBI and characterized 190 putative cathelicidin genes (a family of innate immune peptides), 103 of them previously undescribed [12]. Resequencing studies routinely add public SRA runs to their own data. A sesame genotyping study combined five newly sequenced Egyptian genotypes with 19 SRA datasets to reach 24 genotypes for variant calling and population structure analysis [13]. NCBI search also answers distribution questions. After finding an 0.882 megabase megaplasmid (a very large plasmid) in two Vibrio parahaemolyticus strains, researchers searched NCBI and located the same element in four additional strains spanning three Vibrio species, which reframed it from a curiosity into a rare but globally distributed genetic element [14]. Even phenotype claims get checked against [NCBI genomes](/blog/guides/ncbi-genomes). A biochemical test suggested that Helicobacter species lack the enzyme L-alanine aminopeptidase, and the absence was confirmed by inspecting the available full genomic sequences of Helicobacter on NCBI [15]. ## When to Use NCBI Taxonomy ![Decision path asking whether the workflow uses NCBI tools, whether scope crosses domains, and whether a specialist marker or catalog is needed.](/article-images/illustrations/ncbi-taxonomy-vs-other-databases-backbone-choice.webp) Choosing NCBI taxonomy or other databases is less a fork in the road than an ordering decision. NCBI first, specialists second. Use NCBI taxonomy as your backbone when any of the following holds. Your pipeline reads NCBI data or NCBI-aware tools. BLAST, Kraken2, Centrifuge, the datasets tool, and SRA all speak taxid. A backbone your tools parse natively removes an entire class of mapping bugs. Your scope crosses domains. Viruses, bacteria, archaea, fungi, plants, and animals sit in one tree. No specialist database covers that range. You will deposit data or publish. Sequence submission, SRA metadata, and most journal reporting standards assume NCBI names and taxids. Reviewers can check an NCBI name in seconds. You need strain and virus nodes. NCBI Taxonomy carries nodes below species, which matters for outbreak work and for viruses where species-level names are coarse. You want the names the archives actually use. If your results must reconcile against GenBank records, SRA runs, and RefSeq genomes, the NCBI classification is the one they share. ## When Another Database Is the Better Choice ### Reach for GTDB When Names Must Follow the Genome Tree Prokaryotic taxonomy has formal names, NCBI names, and GTDB names, and they do not always line up. When your claim is about organismal phylogeny rather than archive bookkeeping, classify your genomes with GTDB-Tk and report GTDB names, ideally alongside NCBI names so readers can cross-walk between them. The dairy study treated the two as separate classification systems and analyzed both, which is the honest way to handle the disagreement. ### Reach for BOLD When COI Is Your Marker For eDNA and metabarcoding, BINs are the de facto species baseline. If your marker is COI, anchor your resolution claims to BINs, use GenBank as the sequence source, and state your clustering threshold explicitly, because the threshold is a taxonomic decision, not a technical detail. ### Reach for MGnify When You Need Catalog Consistency If your question lives inside one environment, such as the rumen, and you want genus and species features that were defined under one uniform pipeline, classify against the relevant MGnify catalog [1]. You trade NCBI's breadth for internal consistency, and for quantitative work that trade is often worth it. ### Reach for Curated rRNA Databases When You Have Amplicons For 16S, 18S, and ITS amplicons, curated references such as SILVA and UNITE reduce misassignment compared with raw archive searches. If you use NCBI sources instead, as the Myochromella study did with GenBank and SRA at a 98% threshold [5], justify the threshold in your methods, because reviewers will ask what it means for species delimitation. ## Worked Scenarios ### Scenario 1: Shotgun Metagenomics for Pathogen Surveillance in Cattle Feces Backbone: NCBI. Build a Kraken2 or Centrifuge database from RefSeq, classify reads, and summarize at genus and species levels. Then handle the two known traps. First, plasmid carryover inflated Salmonella calls in the comparison described above, so either use a plasmid-reassigned database or inspect plasmid-associated hits before calling a species positive [7]. Second, database choice changes the feature space, so if the downstream goal is quantitative, classify against MGnify, GTDB, and NCBI and report how the feature sets differ. The dairy study did exactly that and showed the consequences. [Heritability](/knowledge/molecular-biology/heritability-definition-formula-examples) estimates (the fraction of phenotypic variation attributable to genetic effects) for methane ranged from 0.08 to 0.14, and 16S genus features explained 28% of phenotypic variation in methane, more than any other feature type [1]. Feature definition, which starts with the taxonomy database, moved the biology. ### Scenario 2: eDNA Biomonitoring of Fish Backbone: BOLD BINs, sequences from NCBI. A fish eDNA benchmarking study built 22 metabarcodes from the same mitogenomes, compared their clustering output against a BIN-based taxonomic baseline, and found that the clustering threshold, not the clustering method, was the most important factor driving biodiversity estimates. The study proposed optimal thresholds per metabarcode for delineating taxonomic levels [3]. Practically, that means you pick your metabarcode, cluster amplicons de novo, calibrate the threshold against the BIN baseline, and report over-splitting (dividing one taxon into several) and over-merging (combining several taxa into one) rates at your chosen threshold so readers can compare your numbers with theirs. ### Scenario 3: Describing a New Species Backbone: NCBI for the description, BOLD for the barcode if the group is an animal. Search both GenBank and BOLD before you sequence, because the louse fly case shows how a genus can be represented by a single species while its relatives sit unsampled [4]. Generate your own marker sequence, delimit the species with explicit methods, and deposit in both repositories. A Truncatelloidea study combined morphology, COI, 16S, and 28S fragments with species delimitation methods (ABGD, ASAP, and bPTP) and genetic distances to describe a new genus and species [11]. Name the methods and thresholds in your paper, because reviewers will ask. ### Scenario 4: Viral Discovery from Public RNA-seq Backbone: NCBI. A cocoa virome study pulled 109 RNA-seq libraries from nine BioProjects in SRA, ran a pipeline of viral enrichment steps and hybrid assembly, and searched the results against NCBI non-redundant databases, which surfaced ten putative novel viruses and a novel Badnavirus species [2]. The taxonomy tree is your sampling frame here. Browse the host taxon, list its BioProjects, and download the runs, rather than keyword-searching blindly. ## What to Report in Your Methods Reproducibility in taxonomic work is a reporting problem as much as a computational one. Every methods section that touches taxonomy should state the following. 1. The database name and version or release, for example the RefSeq release, the GTDB release, or the BOLD download date. 2. The date you downloaded any files, because NCBI Taxonomy changes continuously. 3. Which taxonomy you used for reporting, whether NCBI, GTDB, both, or a mapping between them. 4. The classifier and the exact database build behind it. 5. Clustering thresholds for any OTU or ASV work, and the reference database they were set against. 6. How plasmids, unclassified reads, and merged taxids were handled. ## Common Mistakes and Limitations 1. Treating GenBank labels as verified identity. Organism names on submitted sequences are provided by submitters, and mislabeled sequences in archives such as GenBank are a documented problem that distorts downstream taxonomy [3]. For anything consequential, verify the label against a curated reference or a type strain. 2. Letting shared plasmids impersonate a species. The Salmonella comparison showed how plasmid sequences inside bacterial RefSeq genomes inflated species-level calls, and how reassigning plasmids to their own taxon changed the answer [7]. If your classifier database bundles plasmids with chromosomes, your species calls inherit that choice. 3. Mixing taxonomies without a mapping. Genus and species sets differ across NCBI, GTDB, and MGnify, and the dairy study treated them as three separate systems for good reason [1]. Pick a primary taxonomy, map the others onto it explicitly, and say which is which in your methods. 4. Assuming every species is already in the database. The nightjar louse fly had no sequences in GenBank or BOLD until a specimen was finally sequenced [4]. Absence of a match is evidence of a coverage gap, not proof that your organism is new, and a genus-level match is not a species identification. 5. Reading NCBI Taxonomy as a phylogeny. The database is a curated classification for organizing records. For phylogenetic claims, build a tree from aligned sequences and say which tree you used. 6. Not pinning versions. NCBI Taxonomy changes daily, GTDB changes by release, and catalogs change on their own schedule. Record the database, the version, and the download date, or your results will not be reproducible. 7. Treating a metagenomic species call as a diagnostic result. In the Salmonella comparison, culture and PCR agreed with each other and disagreed with the uncorrected metagenomic call [7]. A database match is a hypothesis. For clinical or regulatory decisions, confirm with an orthogonal method. 8. Forgetting that taxonomic resolution is an analysis parameter. Genus-level and species-level features are different datasets with different statistical behavior, and the dairy study measured them as separate feature types [1]. Choose the resolution deliberately and state it. ## Frequently Asked Questions ### What is the difference between NCBI taxonomy and other taxonomic databases? NCBI Taxonomy is a curated classification backbone tied to sequence records in GenBank, RefSeq, and SRA, while databases such as GTDB, BOLD, and MGnify build classification from genomes, COI barcodes, or metagenomic catalogs. The practical difference is that the same reads or genomes can receive different names and different genus and species sets depending on which database you query. ### When should I use NCBI taxonomy instead of GTDB? Use NCBI taxonomy when your workflow uses NCBI tools and archives or spans viruses and eukaryotes, and use GTDB when you need prokaryotic names that follow a genome phylogeny. Many prokaryote projects report both name systems side by side. ### Is NCBI taxonomy reliable for species identification? The taxonomy itself is curated, but organism labels on GenBank sequences are submitter provided, and mislabeling is documented. Treat a database match as a hypothesis and confirm consequential calls with an orthogonal method such as culture, PCR, or a curated reference comparison. ### Can I use NCBI taxonomy for COI barcoding and eDNA studies? You can, but BOLD BINs are the standard species baseline for COI, and a fish eDNA benchmark found that the clustering threshold mattered more than the clustering method for biodiversity estimates. Anchor resolution claims to BINs and state your threshold. ### How do I find the taxid for my organism? Search the NCBI Taxonomy Browser, or script it with the E-utilities by querying the taxonomy database with the scientific name. The query for Escherichia coli returns the record for taxid 562, and the same pattern works for any name in the database. ### Do I have to pick one taxonomic database for my whole project? No. Pick a backbone, usually NCBI, pin the versions of every database you touch, and map specialist names such as GTDB names or BINs onto the backbone where your question needs them. Report each database and version in your methods. ## Sources 1. [Comparison of 16S rRNA sequencing and shotgun metagenome sequencing for estimating genotypic and phenotypic parameters of enteric methane emission in dairy cattle.](https://pubmed.ncbi.nlm.nih.gov/42114750/) 2. [Theobroma cacao Virome: Exploring Public RNA-Seq Data for Viral Discovery and Surveillance.](https://pubmed.ncbi.nlm.nih.gov/40431635/) 3. [Benchmarking the Taxonomic Resolution of Fish eDNA Metabarcodes Against COI Barcodes.](https://pubmed.ncbi.nlm.nih.gov/41168956/) 4. [A Molecular Phylogeny of the Nightjar Louse Fly Pseudolynchia garzettae (Rondani, 1979) (Diptera: Hippoboscidae) with a Diagnostic Morphological Description of a Sequenced Specimen.](https://pubmed.ncbi.nlm.nih.gov/41941016/) 5. [Myochromella unveiled: exploring its global distribution through a public database of amplicons.](https://pubmed.ncbi.nlm.nih.gov/40413396/) 6. [Comparison of machine-learning methodologies for accurate diagnosis of sepsis using microarray gene expression data.](https://pubmed.ncbi.nlm.nih.gov/33999966/) 7. [A Cautionary Report for Pathogen Identification Using Shotgun Metagenomics; A Comparison to Aerobic Culture and Polymerase Chain Reaction for Salmonella enterica Identification.](https://pubmed.ncbi.nlm.nih.gov/31736924/) 8. [Establishment and preliminary evaluation of an RT-ERA combined with CRISPR/Cas12a assay for detection of H5 and H7 influenza viruses.](https://pubmed.ncbi.nlm.nih.gov/42818682/) 9. [Characterization of Trueperella pyogenes strains from uterus of cow and an analysis of their phylogenic association with strains from other anatomic locations and hosts.](https://pubmed.ncbi.nlm.nih.gov/41687473/) 10. [Unveiling the Unique Mitogenome Structure of Phylloporus: Implications for Phylogeny and Evolution in Boletaceae.](https://pubmed.ncbi.nlm.nih.gov/41440657/) 11. [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/) 12. [Exploring diversity in avian immune defence: Insights from cathelicidin clusters.](https://pubmed.ncbi.nlm.nih.gov/40127729/) 13. [Exploring the genetic landscape of sesame: whole genome sequencing reveals distinct Egyptian lineages and SNP variability.](https://pubmed.ncbi.nlm.nih.gov/41619030/) 14. [Megaplasmids of the enteropathogenic species Vibrio parahaemolyticus, Vibrio cholerae and Vibrio vulnificus represent a group of novel genetic elements unrelated to other plasmids of Vibrionaceae.](https://pubmed.ncbi.nlm.nih.gov/41360192/) 15. [An evaluation of the Oxoid Biochemical Identification System Campy rapid screening test for Campylobacteraceae and Helicobacter spp.](https://pubmed.ncbi.nlm.nih.gov/19413776/)