Put GFP into pUC19¶

This page follows one cloning job from end to end. The vector is pUC19. The insert is the GFP coding sequence. One command does the design and writes the bench protocol.

Both sequence files ship with the repo, under tests/data/.

Run it¶

pixi run liulab_mbio goldengate plan tests/data/pUC19.dna tests/data/GFP.dna --out docs/examples/pUC19-GFP

It prints one summary line, then the three paths it wrote:

pUC19-GFP: 3347 bp, BbsI, 2 fragments, overhangs ATGA, TGGC, fidelity 100% (measured), checks warn

Read that line first. It is the whole design in one sentence.

Why BbsI¶

The tool ranks enzymes by how many sites they read in the parts. BsaI reads a site in both pUC19 and GFP, and BsmBI reads two in pUC19. Cutting either would cut the parts as well as the ends. BbsI reads none, so nothing has to be changed to use it.

If no enzyme were free, the command would stop and say so. It never quietly edits your sequence to make one fit.

Why checks warn¶

Nine oligos were designed and five carry a warning: two on GC content, two on the GC clamp and one on melting temperature. None fails. The badge names those kinds, and the oligo sheet gives every oligo a verdict of its own.

Each oligo is judged by what it is for. A sequencing primer may be as short as 16 bases, since one of the free primers Genewiz itself offers is that long. A PCR primer needs 18. The GC clamp asks for one to three G or C bases among the last five. No published rule sets that band, so the sheet calls it proposed.

A warning is yours to judge. The design tries every length a primer could have and keeps the one that warns least, so each of these comes from where that primer has to sit.

What comes out¶

Nothing here is written by hand, so the same two input files always give the same three files. That means you can run it again instead of editing the output.

How to read the protocol¶

Open protocol.html in a browser. It is built to be followed at the bench, not filed away.

The top of the page carries eight cards: the vector, the insert, the enzyme, the fragment count, the overhangs, the fidelity, the product and what to select on. Each is a few words. Under them sit four sentences about what the clone should do, then the badges — sites, junctions, primers and one per part. A badge that is not green spells out why beneath it.

Two of those sentences matter before you start:

• GFP goes in on the opposite strand from the lac promoter, so that promoter does not read it.
• Nothing is annotated ahead of GFP to start translation, so the clone is not expected to glow.

That is the design telling you what it is and is not. It reads both from the finished plasmid, so it cannot disagree with the map.

The oligo sheet gives each oligo its own verdict, written as a word rather than a colour: pass, warn or fail. One line under the sheet opens to show which check fired on which oligo, what it measured, and the band it missed. It stays closed, so the sheet you order from is still a list of oligos.

The middle is eleven numbered steps, from the first PCR to sequencing. Each step has a check mark, what to do, what a good result looks like, and what to do when it is wrong. Reaction tables rescale when you change the number of reactions at the top of the table, so you do not do that arithmetic yourself. Gel steps draw the bands to expect.

The screening step is the one to read twice. A correct clone gives two bands, 160 and 897 bp. An empty vector gives one band at 236 bp. A clone with GFP the wrong way round gives 220 and 897 bp. Those three lanes are different on purpose, so the gel alone tells you which you have.

The finished example¶

The plan above is published here, exactly as the command wrote it:

• the bench protocol
• everything the run wrote

Point it at your own ligase data¶

Fidelity says how likely the joins are to pair correctly. For five enzymes that number is measured, and the measurements ship with the package. For the rest, it is worked out from rules.

You can do better if you hold a copy of the T4 ligase data from Potapov and colleagues. That archive is not ours to ship, so nothing from it is in this package. Point the command at your own copy instead:

pixi run liulab_mbio goldengate plan vector.dna insert.dna --out plan/ \
    --ligase-matrix ~/potapov/FileS03_T4_18h_25C.xlsx

An environment variable does the same for every run:

export LIULAB_MBIO_LIGASE_MATRIX=~/potapov/FileS03_T4_18h_25C.xlsx

Reach for FileS03 first: 25 °C for 18 hours is what NEB's own viewer assumes. Both .xlsx and .csv are read.

The report then says where its number came from, and keeps the three kinds apart:

The enzyme's own measurements win where they exist. A ligase file stands in for a measurement nobody has made; it does not replace one somebody has. Without such a file nothing changes.

Before you order¶

• Read the whole protocol once, top to bottom.
• Look at every badge that is not green, and decide about it.
• Open the line under the oligo sheet and decide about each warning.
• Check that the product map holds what you meant to clone.
• Check the three screening bands are different from each other.

pixi run liulab_mbio goldengate plan --help lists the rest of the options: where to put the insert, which way round it goes, holding a join in frame, forcing an enzyme, and which polymerase and strain the protocol should name.