unsloppable

Ten thousand clones.One line.

We transfect, deposit single cells, image every well from day zero, and hand back one clone with its record.

Start a line
The programme
01/ 03

A cell line used to take eleven months and four people, and the clone you shipped was the one that happened to grow, not the one that was best.

We deposit one cell to a well, photograph it on the hour from the day it lands, and keep every frame, so monoclonality is something you can look at, not something you assert.

What leaves here is one line, its titer curve, sixty passages of stability data, and the image record that proves, well by well, where it came from.

The work01

We do not pick clones. We rank them.

Twelve thousand clones come off the deposition head each week. Every one is photographed, sized and scored before it is two days old, so what you choose from is a ranked list rather than a plate somebody liked the look of.

See the ranking

No hand touches a plate between transfection and pick.

01.

One cell per well

The deposition head places one cell in one well and photographs the well it landed in.

02.

Imaged on the hour

Every well is imaged hourly for twenty-one days. Nothing is sampled, and no frame is discarded.

03.

Ranked before you look

Growth rate, titer and stability score every clone, so the shortlist exists before anyone opens the incubator.

01
02
03
04
05
06
07
08
09
Stage map, run 4,182

Nine stages, start to bank.

Stages 01 to 04

Transfection, selection, deposition, imaging. Four weeks with nobody at the bench, and the first ranked list at day twenty-eight, ordered by titer.

Stages 05 to 09

Expansion, stability, banking, release. The line you picked at week four is still the line at passage sixty, or you hear why.

The image record860,000 frames a day
Published work

What we have published.

09 / 09

  1. 012026Hourly imaging of 384-well plates for twenty-one daysWhat a clone looks like on day two, and how well that predicts day sixty.Cell Eng. Lett.
  2. 022026Monoclonality from the image record, not from statisticsTwo consecutive frames of an empty well and a single cell, kept for the file.Clonal Reports
  3. 032025Titer at passage sixty against titer at passage fiveThree hundred lines followed to the end of stability. Two thirds held within ten percent.Bioprocess Rev.
  4. 042025Single-cell deposition at ninety-eight percent first passThe head, the settling time, and why the last two percent are worth chasing.Assay Methods
  5. 052025Segmenting colonies that touch the well wallThe failure mode that made every early growth curve optimistic, and the fix.Cell Eng. Lett.
  6. 062024A titer curve is a better clone score than a titerRanking on the shape of the curve moved the top clone eleven places on average.Bioprocess Rev.
  7. 072024What the pool already tells you about the clonePool titer explains less of the final result than anyone on the floor expected.Yield Notes
  8. 082023Six hundred million well images, and what we keepStorage, retention and the argument for never discarding a frame from a run.Culture Index
  9. 092023Stability without a selection agent, out to passage ninetyRemoving pressure at passage ten cost four percent of titer and nothing else.Assay Methods

What a line costs and what leaves with it.

Screen$148,000Per line

Transfection, deposition, imaging, and a ranked list.

Develop$265,000Per line

Screen, plus expansion and stability to passage sixty.

Transfer$410,000Per line

The line, the bank, the image record, the dossier.

Ranked clone list
Titer curve to passage sixty
Stability data and image record
Research and master cell bank
Clonality dossier for filing
A programme is four lines run together on the same construct.Take a slot

Slots are quoted per construct and confirmed after a feasibility read. Priced by the line, never by the well.

From the floor.

All notes
Run 4,182

Method2026.02

Why we kept every frame of run 4,182

Six hundred million images is an expensive habit. Then a customer asked which well a line came from in 2023, and we could answer in four minutes.

Method2026.01

A settling time nobody had measured

Programme2025.11

Sixty-one lines banked this year

Note2025.09

The incubator door is the variable

Next intakeFour slots open

Send us the construct. We will send back the line.Send a construct