Reformatting looks like a bookkeeping exercise. You have samples in one plate format and you want them in another, and the deck can clearly reach every well, so the whole thing feels like a matter of getting the map right. That view will get you a run that finishes without an error and a set of results you cannot trust, because the part that actually matters is not the map. It is that the well you are dispensing into has changed shape, changed volume, and changed how liquid behaves at its bottom, and none of that is captured by a spreadsheet of source and destination positions.
So treat reformatting as two problems wearing one name. The first is the geometry of the transfer: which source well feeds which destination well, and how the head moves to service it. The second is the fluid physics of the destination: whether the liquid class that worked in the plate you are leaving still works in the plate you are arriving at. The first problem is easy to see and easy to get right. The second is the one that quietly ruins plates, and it is where most of this article lives.
The three moves you will actually make
Almost every reformatting job is a variation on one of three patterns, and it helps to name them before talking about what each one costs.
- Consolidation: you collapse four 96-well plates into a single 384-well plate, interleaving them by quadrant so that plate one fills the wells at even rows and columns, plate two shifts one column over, and so on until the four source maps tile cleanly into the denser destination.
- Expansion: you run consolidation in reverse, fanning one 384-well plate back out into four 96-well plates, usually because a downstream instrument or assay only speaks the lower density.
- Replication and stamping: you copy a plate to one or more plates of the same format, one-to-one, which is the simplest case geometrically but still inherits every fluid-physics concern below the moment volumes are small.
Consolidation is where the interleaving matters most. Four 96-well plates map into a 384 by quadrant: think of the 384 as four interlaced 96-well grids offset by one well in each direction. Get that offset wrong by a single column and every sample lands in the wrong place, silently, because the run still completes. That is the classic reformatting failure, and it is worth a deliberate check of the first and last wells of each quadrant before you trust a new map.
What changes when the well shrinks
Here is the thing that catches people. A liquid class validated at 200 microliters in a 96-well plate is not the same liquid class at 20 microliters in a 384-well plate, and reusing it because it shares a name is how you inherit errors you never measured. The smaller well changes almost every assumption the class was tuned against.
- Working volume: a 384-well holds a fraction of what a 96-well holds, so you are operating at the low end of the tip range where relative error is largest and every settling and blowout parameter matters more.
- Well geometry and meniscus: a narrower well means a tighter meniscus and a sharper relationship between volume and liquid height, so a small volume error shows up as a large height error.
- Immersion depth: there is simply less column of liquid to work with, so the safe window between submerging the tip and hitting the bottom or breaking the surface is far narrower than in the plate you left.
- Liquid-level detection sensitivity: with a shorter liquid column and a smaller cross-section, level sensing has less signal to work with and is easier to fool, so a threshold that was comfortable at 96 can become marginal at 384.
- Dead volume: the unusable liquid at the bottom of a well is a much larger fraction of a small well, so consolidation can leave you short even when the arithmetic said there was plenty.
None of these are exotic. They are the ordinary consequences of a smaller vessel, and each one pushes in the same direction: the denser plate is less forgiving, so the class needs to be revalidated at the new volume range rather than assumed to carry over. Run the gravimetric check at the volume you will actually use in the 384, not the one you validated in the 96.
The head has to move differently too
Reformatting across densities is also a mechanical change, because the tip pitch that lands cleanly on a 96-well grid does not land on a 384-well grid. Servicing every well of the denser plate means either a head that can address the tighter pitch or a sequence of offset passes, each shifted by a quadrant, so the same tips visit the interleaved positions in turn. Those offset moves are extra opportunities to mismap, which loops straight back to the check you already owe on quadrant alignment.
The denser plate also brings the wells physically closer together, and that geometry has a fluid cost. Splashing that would have landed harmlessly on an empty part of a 96-well plate can now reach a neighbouring well, and any aerosol or droplet thrown during a fast dispense has a shorter distance to travel to cause cross-well contamination. This is the trade at the heart of the format choice: the 384 buys you throughput and reagent economy, and it charges you in contamination and splash risk that a slower, gentler dispense has to pay back down.
Where reformatting quietly fails
Most reformatting disasters are one of a small set, and all of them survive the run without raising an error, which is exactly what makes them dangerous.
- Wrong quadrant mapping: an offset applied in the wrong direction or by the wrong amount ships every sample to the wrong well, and nothing downstream knows until the results make no sense.
- Cross-well contamination: a dispense tuned for the roomier plate throws droplets into neighbours once the wells crowd together, so a class that passed at low density fails at high density for reasons that have nothing to do with volume accuracy.
- Splashing: a fast plunge or an aggressive blowout at the bottom of a shallow, narrow well ejects liquid upward, costing you volume and seeding contamination at the same time.
Reformatting is not moving liquid to new addresses. It is asking whether the liquid class you trusted in one vessel is still true in a smaller one, and the honest answer is only known after you measure it.