Cloning a Gene From Scratch: A Complete Workflow From Sequence to Verified Construct

Step 1 — Fetch the exact source sequence

Fetch the insert and vector sequence by accession or record rather than retyping a plasmid map from a paper or copying a sequence out of a supplementary figure. A single transcription error here, a dropped base, a transposed pair of letters, is invisible to every downstream check that only looks at what you typed, and will not surface until a primer that should anneal cleanly does not, a junction that should assemble does not line up, or a Sanger trace comes back with a mismatch that is actually just your own typo. Fetching directly from the source record also pins the exact version, which matters if the record is later revised.

The fetch-a-sequence-by-accession-number guide covers this in depth; the short version here is to start both insert and vector from a fetched record, not a retyped one. Sequence Fetcher does exactly this step: paste in a GenBank, RefSeq or UniProt accession and get back the FASTA or GenBank record to carry into every step that follows.

Step 2 — Choose an assembly strategy before designing primers

Restriction and ligation cloning, Gibson Assembly, and Golden Gate Assembly all join fragments in a defined order, but each needs something different from your primers, matched sticky ends, homology arms, or defined Type IIS overhangs, so the choice has to come before primer design, not after. Which one fits mostly comes down to fragment count and whether the parts need reuse in other builds; a one- or two-fragment construct with good existing sites is a different problem from a five-part combinatorial library.

The tradeoffs are covered in full by the Gibson Assembly vs. Golden Gate vs. restriction cloning guide; the point here is to lock the method in first, since redesigning primers after picking the wrong strategy midway through is wasted, avoidable work.

Step 3 — Design the primers or homology arms

With a method chosen, design the actual oligos:

• Match melting temperature (Tm) between the forward and reverse primer in a pair so both halves anneal at the same step; a several-degree mismatch is a common reason an otherwise sound reaction fails.
• For Gibson-style builds, give each junction 15-40 bp of homology exactly matching the neighboring fragment; a single base out of place at either end of that overlap can keep the assembly chemistry from finding a clean joint.
• Before ordering anything, check each primer's specificity against the template with an in silico PCR-style check: does it also anneal somewhere else in the vector or insert, producing an unintended second product.

Step 4 — Assemble the construct in silico before ordering anything

Build the intended assembled sequence on paper before a single oligo is synthesized. This is the step that catches a homology arm one base short of a true match, a Type IIS overhang that does not actually pair with its intended neighbor, or a restriction or Type IIS site sitting inside an insert that nobody screened for, any of which looks fine on paper but will not assemble, or will assemble wrong, on the bench. Catching it here costs a few minutes; catching it after a failed transformation costs the whole reaction.

Cloning Simulator does this assembly step directly: give it the fragments for a Gibson, Golden Gate, or restriction cloning build and it assembles them and designs the junction primers needed, so you see the finished sequence and every junction before ordering anything.

Step 5 — QC the assembled construct, not just the parts

Checking each fragment individually is not the same as checking what they become once joined. The assembled construct needs its own pass: reading frame has to stay correct across every junction, which matters most where a fusion tag joins a coding sequence, since a junction off by one base shifts the frame for everything downstream of it. Scan the assembled sequence for premature stop codons introduced at a joint, cryptic ribosome-binding-site-like motifs or polyA-like signals absent from either fragment alone, restriction sites created only by the new junction sequence, and GC extremes concentrated right at a joint. None of these show up when you only inspect the original fragments; they are artifacts of stitching sequence together.

Construct QC Linter runs this scan on the assembled coding sequence and flags exactly this class of problem before you commit the design to a real reaction.

Step 6 — Screen colonies, then verify by sequencing

After transformation, colonies get screened first, typically by colony PCR, to find the ones carrying an insert of the right size, a useful first pass, though one with its own failure modes around primer placement and lysis efficiency worth troubleshooting separately. The candidates that pass move to Sanger sequencing against the expected reference, and that comparison deserves more care than eyeballing the trace: the read can come off either the forward or reverse primer, so check orientation before assuming a mismatch is real, and trim the low-quality bases at both ends before trusting anything there.

Reading a chromatogram and telling a real mutation from sequencing noise is its own guide; in short, align the trimmed, correctly oriented read against your reference and trust only mismatches sitting in clean, high-quality trace. Sanger vs Reference does this alignment directly: feed it the Sanger read and the expected reference sequence and it returns a pass or needs-review verification report instead of a trace you have to check base by base.

Where most cloning projects actually go wrong

Most cloning failures trace back to one of a small number of checkpoints, and catching each one in silico, at the specific step above where it belongs, is cheaper than debugging a failed transformation after the fact.

• A Tm mismatch between a primer pair, caught at the primer-design step.
• An internal restriction or Type IIS site nobody screened for, caught at the in silico assembly step.
• A homology arm shorter than assumed, caught the same way, before it costs a reaction.

Related references

Related tools

Related guides

• Gibson Assembly vs. Golden Gate vs. Restriction Cloning: Choosing a Strategy →
• How to Design Site-Directed Mutagenesis Primers (QuikChange and Q5-Style) →
• How to Verify a Clone with Sanger Sequencing →