So the complexity of a rolling door, like a garage door, is that it must be broken into segments, with each segment being a "leak point" for any leaking air.
The most possible configuration today would involve using air pressure to seal the gaps.
Basically the rolling door would actually be rather loose around the guide rail, like a garage door. Each slat would have a bit of play to make it easy to roll up and roll down. In a vacuum, this door would simply operate like a garage door in atmosphere.
Once the door was closed, however, air would be directed into the air lock. As pressure increases, the slats of the door would be "blown" into the door frame; gaskets along the edges of the slats would seal against the frame and each other.
If you want a more secure joint, or something with more control, use magnets.
This door still depends on internal pressure to keep the door closed, but uses electromagnets to properly align the slats during the low pressure phase of the door closing. I'd expect there to be a V shaped notch in the rail to "seat" the slat in the proper position. During vacuum operation, the slats only use the rails to keep them relatively in place as they are being rolled up or down.
When it's time to seal the door, the slats are rolled down. Just before each slat rolls into it's "home" position (where the notches are), electromagnets are engaged. This pulls the slat into the "notch", ensuring that the slats remain seated in place. Again, gaskets are used to "seal" the slats against each other and the door frame. This has the benefit that no air leaks during the re-pressurization phase, as the doors are "held closed" by the magnets until the air pressure is great enough to do the job itself. Then the magnets can be shut down.
Most early space vehicles today didn't run at 1ATM, so you could choose to run a thinner atmosphere mix. An aluminum or synthetic sheet, with ribs to provide structural support, might be "good enough" to keep in air pressure, instead of "slats" in above. Note that there are a TON of reasons why we now use 1ATM on our space vehicles, essentially involving mixing atmospheres while docking and the need to avoid "the bends" when transitioning to an Earth style environment. Your decision on if the added bulk to support a 1ATM Earth-like atmosphere is valid in your universe (as it was an assumption of your question.) Also, you COULD run your shuttlebay at 1/3 of an atmosphere (about the top of mount everest), if the shuttles themselves were fully pressurized and docked onto a pressurized port. This would protect from explosive decompression (the crew could survive the decompression) of ships and space suits while repairing the ship. There's a lot more engineering to think about doing this, however.
If you don't mind the door being a bit ridiculous... you could use a large balloon, maybe. I'm not an engineer, but I imagine a large balloon, which can be pressurized to 2ATM. When it's time for the door to close, the balloon is inflated; it's slightly larger than the door it's sealing off, so the balloon "seats itself" against the door frame. Guide wires could be on tracks inside the frame to ensure the balloon actually seats against the door. Because there is vacuum on both sides of the balloon at this point, it would grow in both sides, seating itself against the round aperture. Once the seal was made, the cabin could then be pressurized.
The balloon itself would use it's higher-than-ambient pressures to "seal" against the portal, which would protect the internal pressure from escaping into space. Because the balloon is higher-than-ambient pressure, internal pressures would not force the balloon out into space. I think this is a silly idea, but it might work.