Biological databases: NCBI, UniProt, Ensembl

A biological database is an organized online collection of biology data, such as DNA sequences, proteins, and research papers, that anyone can search. The largest ones are free and run in an ordinary web browser, so you do not need to write any code to use them.

NCBI: the central hub

NCBI, the National Center for Biotechnology Information, is a United States government website that gathers many resources in one place. One of them, GenBank, stores nucleotide sequences: the long strings of letters (A, C, G, and T) that spell out DNA. GenBank is open for submission, which means researchers around the world upload their own sequences to it.

RefSeq is NCBI's curated set of reference sequences, meaning experts review and tidy them into one clean, official record per gene or protein. PubMed is NCBI's search tool for summaries of published research papers, called abstracts. Because these resources live together, a single trip to NCBI can hand you a sequence and the studies written about it.

UniProt for proteins

A protein is a tiny molecular machine that a gene's instructions build, and proteins do most of the work inside living cells. UniProt is the go-to protein database: it gives each protein its amino-acid sequence (the letters that make up a protein) plus detailed functional annotation, which describes what the protein does, where in the cell it acts, and how it connects to other data.

Ensembl genome browser

A genome is the complete set of DNA in an organism. Ensembl is a genome browser, a website that lets you scroll and zoom along a genome to see where each gene sits, much like panning across a map. It hosts annotated genomes, meaning the sequence is labeled with the locations of genes and other features, for humans and many other species.

Accession numbers

An accession number is a stable, permanent ID for a single database record, so you can always find the exact same entry again and cite it. The form of the ID hints at its source: RefSeq messenger RNA (a working copy of a gene) records begin with NM_, for example NM_007294, while a UniProt protein record uses a short mix of letters and numbers such as P38398. Submitted data like GenBank is primary and raw and can overlap or contain small errors, whereas curated data like RefSeq has been checked and cleaned by experts.