How to Identify an Unknown Sequence: DNA, RNA or Protein
6 min read · Updated July 10, 2026
Before you can figure out which gene, organism, or protein a sequence belongs to, you sometimes have to answer a more basic question: is this even DNA, RNA, or protein. A sequence pasted from an email, a lab notebook, or an unlabeled file can arrive with no indication of molecule type, and feeding it into the wrong tool — a codon table for a peptide, a nucleotide search for a protein — wastes time or returns nothing useful. This guide covers how to read the character alphabet to make that call, why the check gets unreliable for short sequences, and how to move from knowing the molecule type to actually identifying the sequence with a database search.
Start with the character alphabet
The cheapest signal for molecule type is which characters the sequence actually uses. Nucleic acids are built from a four-letter core, while protein sequences are written in the one-letter amino acid code, so scanning the letters usually settles the question before you need any other analysis.
- DNA — only A, C, G, T, plus an occasional N (unknown base) or other IUPAC ambiguity code such as R, Y, W, S, K, or M.
- RNA — the same four-letter core, but with U in place of T.
- Protein — drawn from a roughly 20-letter amino acid alphabet, including letters such as E, F, I, L, P, and Q that never appear in a nucleic acid sequence.
Why a few letters usually settle it
That last point is normally the fastest tell. Because so many amino acid letters fall outside the nucleic acid alphabet, finding even one E, F, I, L, P, or Q anywhere in a sequence is enough to know you're looking at a protein — you don't need to read the whole thing, just enough of it to hit a letter that ACGTUN can't explain.
Where the alphabet check breaks down
Composition is a strong heuristic, not a proof, and it gets weaker as a sequence gets shorter. The catch is that four amino acids share their one-letter codes with nucleotide letters: alanine is A, cysteine is C, glycine is G, and threonine is T. A sequence built entirely from A, C, G, and T is almost certainly DNA once it runs a few dozen bases long, but a short fragment of only six or eight characters drawn from that same set could just as validly be read as a short peptide made of alanine, cysteine, glycine, and threonine residues. There's no way to resolve that ambiguity from the letters alone — trust the alphabet check at realistic sequence lengths, and treat it as inconclusive for very short queries.
Composition tells you what, not who
Knowing that a sequence is DNA, RNA, or protein only answers the molecule-type question. It doesn't tell you which gene, which organism, or which specific protein you're holding — that requires comparing the sequence against a database of known sequences. A BLAST search does exactly that: it searches public sequence databases for regions of similarity and returns a ranked list of hits, each scored by how closely it matches your query. Instead of guessing at identity from composition or length, you get the closest known matches, with alignments you can inspect to decide which hit actually explains your sequence.
A practical workflow
In practice, the two checks are complementary and run in order, not as alternatives.
- Check the alphabet first. This confirms molecule type and catches simple mistakes, such as accidentally pasting a protein sequence into a tool that expects DNA, before you waste time on the wrong analysis.
- Once you know the type, hand the sequence to BLAST to find out what it actually is, or route it to a type-appropriate analysis tool — ORF and restriction-site analysis for DNA, composition and motif analysis for protein — once identity or type is established.
Let SeqBench check the alphabet and search in one step
Characterize Sequence runs this workflow for you: paste any unknown sequence and it produces a full identity report — molecule type and composition, with the alphabet check already done — and lets you send the sequence straight to BLAST from the same page. That removes the manual letter-scanning step and the copy-paste into a separate search tool. If you already know the sequence is DNA and want composition, ORF, restriction-site, or primer information instead, Sequence Analyzer covers that from the same pasted sequence. If you're starting from an accession number rather than raw letters, Sequence Fetcher pulls the FASTA or GenBank record from GenBank, RefSeq, or UniProt so you have an actual sequence to check in the first place.
Frequently asked questions
How can I tell if a sequence is DNA or protein just by looking at it?
Check which letters appear. If the sequence uses only A, C, G, and T (plus maybe N), it's DNA; if it also contains letters like E, F, I, L, P, or Q, which never appear in DNA or RNA, it's protein.
What's the difference between a DNA and RNA sequence alphabet?
They share the same four-letter core, but DNA uses T (thymine) where RNA uses U (uracil) — everything else about the letters is identical.
Can a short sequence be read as either DNA or protein?
Yes. Alanine, cysteine, glycine, and threonine all share their one-letter amino acid codes with nucleotide letters (A, C, G, T), so a short fragment made only of those four characters can't be definitively typed by composition alone.
When should I BLAST a sequence instead of just checking its letters?
As soon as you need to know what the sequence actually is — which gene, organism, or protein — rather than just what kind of molecule it is. The alphabet check only tells you molecule type, not identity.
Related references
Common restriction enzymes: recognition sites, cut positions, NEB buffer activity, star activity and an interactive double-digest buffer finder.
Quick reference for FASTA, FASTQ, GenBank and related formats.
PCR primer design rules for length, GC content and Tm.