Reverse Complement and Strand Operations

Reverse-complement is the single most error-prone operation in genomics scripting. Get the orientation wrong and your gene is on the opposite strand, your primer doesn't match, your probe is a hairpin. This skill gives you the right call for every case: complement vs reverse-complement, IUPAC ambiguity handling, batch processing, and the" "which strand am I on?" sanity checks.

When to use

• Computing reverse complements of DNA/RNA for primer/probe design.
• Translating minus-strand genes.
• Converting a feature's coordinates from one strand to the other.
• Building batch reverse-complement pipelines for FASTA/FASTQ.

When NOT to use

• You actually want just the complement (preserving 5'→3' direction) — use .complement().
• For very high-throughput reverse-complement (millions of reads), use seqkit or pyfastx.

Prerequisites

• biopython>=1.83
• For shell-scale: seqkit seq -r (reverse) and seqkit seq -p (reverse-complement)

Core workflow

1. Decide: complement or reverse-complement?
2. Use Bio.Seq's built-in methods (which handle IUPAC ambiguity correctly).
3. For batch, stream records through a generator.
4. Always log a sanity check (e.g., reverse-complement of reverse-complement is the original).

The orientation decision

Code patterns

Basic reverse complement

from Bio.Seq import Seq

s = Seq("ATGGCCATTGTAATGGGCCGCTGAAAGGGTGCCCGATAG")
s.reverse_complement()
# Seq('CTATCGGGCACCCTTTCAGCGGCCCATTACAATGGCCAT')

Complement (no reversal)

s = Seq("ATGC")
s.complement()   # Seq('TACG') — still 5'→3', just base-paired

IUPAC ambiguity is built-in

s = Seq("ATGNRYWSKM")
s.reverse_complement()
# Seq('MKRSWYNCAT') — all ambiguity codes are handled

IUPAC complement table:

RNA reverse complement

from Bio.Seq import Seq

rna = Seq("AUGCCAUUG")  # 5' → 3' mRNA
rna.reverse_complement_rna()  # 5' → 3' of the antisense strand

Or transcribe to DNA first:

dna = rna.back_transcribe()
dna.reverse_complement()

Reverse-complement every record in a FASTA

from Bio import SeqIO

def rc_records(in_path, out_path):
    with open(out_path, "w") as out:
        for rec in SeqIO.parse(in_path, "fasta"):
            rec.seq = rec.seq.reverse_complement()
            rec.id = f"{rec.id}_rc"
            rec.description = "reverse complement"
            SeqIO.write(rec, out, "fasta")

rc_records("input.fasta", "rc.fasta")

Batch with Polars/streaming

from Bio import SeqIO

with open("rc.fasta", "w") as out:
    for rec in SeqIO.parse("primers.fasta", "fasta"):
        rec.seq = rec.seq.reverse_complement()
        SeqIO.write(rec, out, "fasta")

Sanity check: revcomp of revcomp is identity

def assert_revcomp_identity(s):
    assert s == s.reverse_complement().reverse_complement()

Extract a feature and reverse-complement

from Bio import SeqIO

rec = next(SeqIO.parse("chr1.gb", "genbank"))
for feat in rec.features:
    if feat.type == "CDS" and feat.location.strand == -1:
        nt = feat.extract(rec.seq)
        cds = nt.reverse_complement()  # now on + strand
        print(cds.translate())

Reverse-complement a region of a chromosome

def rc_region(record, start, end):
    return record.seq[start:end].reverse_complement()

Common pitfalls

• Complement instead of reverse-complement. The most common bug. If you're on the minus strand and you call .complement(), you'll be working on the wrong strand.
• Lowercase sequences. .reverse_complement() works on lowercase; just remember the output is whatever case the input was.
• RNA vs DNA. reverse_complement() is for DNA. For RNA, use reverse_complement_rna() or transcribe first.
• Strand annotation lost. If you take feature.extract(record.seq), the extracted Seq is a fresh sequence and the feature's strand info is gone. Reverse-complement explicitly when you need to flip strand.
• Assuming IUPAC ambiguity in hand-rolled code. Don't write your own s.translate(str.maketrans("ACGT", "TGCA")) — it breaks on N, R, Y, etc.

Validation

• s.reverse_complement().reverse_complement() == s.
• The first base of the reverse-complement is the complement of the last base of the input.
• Length is preserved.
• GC content is preserved (gc_fraction of input and rc are equal).

Open alternatives

References

• Biopython reverse-complement: https://biopython.org/docs/latest/api/Bio.Seq.html
• IUPAC codes: https://www.bioinformatics.org/sms/iupac.html
• Companion: ors-bioinformatics-sequence-seq-objects, ors-bioinformatics-sequence-transcription-translation.

Changelog

• 1.0.0 (2026-06-10): Initial adaptation by Pradyumna Jayaram from bio-reverse-complement (bioSkills-main/sequence-manipulation/reverse-complement).