I think we know what the question is and I think
@MelbME would agree. It is whether any skewing to VH3-30 is occurring because of a shift in checkpoint-dependent selection in the bone marrow environment or once the cells have gone out into lymph nodes. If we could answer that it would be of huge value.
But it needs very careful usage of controls and normalisations of data. If bone marrow sampling was easy it would be a viable project - maybe a pilot study of 50 pro bands and 50 controls and then scale up and replicate with 200 of each. But that would be very tough.
It is not so much looking for a needle in a haystack as trying to establish whether heartsease is a commoner weed in Martlesham fields than Boromeswell fields using 20 samples of ground ten metres in diameter at each site when heartsease grows in scattered clumps depending on the amount of moisture, sunshine, slope, recent fertiliser dressing and what crops were grown last year. You might need 1000 samples, not because heartsease is hard to find but because it is patchy.
I suspect that to get a reliable estimate of a B cell repertoire shift you might need ten bone marrow samples from the same patient.