What a Double-strand Donor Carries That an Oligo Cannot

A dsDNA donor versus ssODN for knock-in is a template-class split, not a pathway name. Homology-directed repair needs a donor molecule: a short single-stranded oligo or a longer double-stranded fragment or plasmid. Prefer an ssODN for a small substitution that fits synthesis. Prefer dsDNA when the payload is a cassette an oligo cannot carry. Reject “we are doing HDR, so the donor is the same.”

NHEJ versus HDR still owns the pathway names. Knockout versus knock-in still owns those two edit jobs. Richardson et al., 2016 treat donor DNA design, including asymmetric single-stranded donors, as a published class. Paper-specific rates stay in that paper. They are not an official Zetta efficiency. A knock-in that needs HDR still has to choose a molecule. Pathway choice does not fill the order form.

A Long Cassette Usually Needs dsDNA; a Codon Swap May Not

The search compares a dsDNA donor with an ssODN for knock-in. The entities are molecules. An ssODN is a synthesized single strand with homology arms and a small central edit. A dsDNA donor is a double-stranded fragment or a circular plasmid with longer arms and, often, a selectable or fluorescent cassette. Length and chemistry travel together. Oligo synthesis has a practical length ceiling. A multi-kilobase insert does not become an oligo because someone typed “ssODN” on a slide.

dsDNA donors introduce their own constraints. Linear dsDNA can be a substrate for other repair outcomes. A plasmid donor can persist as extra-chromosomal DNA and complicate screening. Those are reasons to design carefully. They are not reasons to order an oligo for a cargo the oligo cannot hold. Do not invent a crossover length as an official rule. Ask what payload you need, then ask whether synthesis or cloning can physically make that donor.

What Each Template Class Can Carry

Score both classes on payload, not on a 2026 HDR rank.

Prefer an ssODN when the intended change fits synthesis and you want to avoid building a donor plasmid. Prefer dsDNA when the payload is a cassette or a replacement that exceeds a responsible oligo. Reject an ssODN order for a fluorescent protein fusion that was never going to fit. Reject a dsDNA plasmid donor if the only intended change is one codon and the lab has no plan to screen out residual plasmid. Published asymmetric-oligo designs can improve some small HDR events under their reported conditions. That is a method paper, not a guarantee for your cell line.

PAM in CRISPR-Cas9 still owns the motif the nuclease needs. A donor class does not replace a PAM. Cut site and donor homology still have to be designed together after the molecule class is named.

This Is Not the Pathway Page

If the reader is asking whether the cell will use NHEJ or HDR, stop and use the pathway sibling. If the reader is asking whether the experiment is a knockout or a knock-in, use that sibling. Stay here only for the donor molecule. A knockout that relies on NHEJ often needs no donor. A knock-in that relies on HDR always needs a donor class. Those sentences are boundaries, not a rewrite of the pathway essay.

Do not claim a Zetta HDR product. Zetta does not ship oligos or donor plasmids. Golden Gate is not a Zetta feature and is not a donor class.

A worked refusal helps more than a slogan. A lab wants a GFP knock-in at a safe-harbor locus. The payload is a full fluorescent cassette plus homology arms. That cargo is a dsDNA-donor class: clone it, confirm the junctions, and plan a screen that can tell integrated cassette from leftover plasmid. Ordering an ssODN because “HDR donors are oligos” fails before transfection. The opposite refusal is also useful. A lab wants one codon change and no extra sequence. Building a 6 kb donor plasmid is allowed, but it is the heavier class. An ssODN, designed to the cut and screened across a small window, matches the payload. Neither example is a promised HDR rate. Each example is a payload that fits one class and not the other.

Name the Spacer After the Donor Class

Name the donor class before anyone orders DNA. After the class is named, a design surface such as ZettaCRISPR can show a named spacer and related sequencing-primer work for the edit you intend. That is a place to hold the guide after the molecule class exists. It does not synthesize the donor, does not promise an HDR rate, and is not the subject of the comparison.

If the order form still says “HDR donor” without ssODN or dsDNA, you do not have a template. You have a pathway word. Write the class. Then order. Then screen the junctions that class can actually create.

Screening follows the class. An ssODN edit is usually scored in a short window around the cut: the intended codon, a silent PAM or restriction change if one was designed, and no unexpected indel. A cassette donor is scored at both junctions and, if a plasmid was used, with a test that residual circular DNA cannot explain a PCR band. Those screens are not interchangeable. A short-window Sanger of a GFP knock-in does not prove the cassette is in. A junction PCR of a one-codon edit does not prove the codon. Write the screen when you write the class. Do not invent an official Zetta HDR percentage to close either screen.

Homology-arm length is a design detail inside the class, not a third class. An ssODN still has short arms around a small edit. A dsDNA donor still has longer arms that must be cloned or synthesized. Changing the arm length does not turn a cassette into an oligo. Do not treat “we lengthened the homology” as a reason to keep an ssODN order for cargo the oligo cannot hold.

Frequently Asked Questions

Can an ssODN carry a long fluorescent cassette?

Usually no. Long inserts are a dsDNA-donor class. An oligo has a synthesis ceiling the cassette does not care about.

Does picking HDR decide dsDNA versus ssODN?

No. HDR is the pathway. Donor class is the molecule you order. The pathway name does not fill the payload.