The NIST source preparation material discusses how RM8398 was extracted from the NA12878 cell. It notes that the cell lines are subject to mutation and so the sequence can start to vary over time. NIST controls for this by producing a large batch of NA12878 and mixing the extracted DNA to reduce the variation between the pooled DNA sequences (at least that's what I understand from the notes):
SOURCE PREPARATION ( 1)Coriell Institute for Medical Research(Camden,
NJ) grew a large growth of their cell line GM12878 in multiple stages,
produced approximately 83 mg of extracted DNA, and then mixed the DNA
and aliquoted it into vials, with the DNA divided approximately
equally into vials. Specifically, the pool of cells was split into
three separate volumes for DNA extraction, and the extracted DNA was
re-pooled and gently mixed at 4 °C for greater than 48 h before the
material was aliquoted automatically into vials of 10 μg of DNA. Note:
This RM is isolated DNA rather than live cells because cells
are less stable and can mutate with each cell division, so
that the sequence of live cells may not be stable over time.
Extracting DNA from a large batch of cells helps ensure that all vials
contain essentially the same sequences of DNA. DNA is currently
available from this same cell line from Coriell with the number
NA12878, but it may contain small differences in the DNA sequence due
to different mutations occurring in different batches of the cells.
Stability:
Stability was assessed by measuring the size distribution of DNA with pulsed field gel
electrophoresis (PFGE). Using PFGE, no change in the size
distribution was detected after storage at 4 °C for eight weeks, but
the size distribution decreased significantly when stored at 37 °C for
8 weeks. In addition, no change was detected after five freeze-thaw
cycles, pipetting vigorously, or vortexing. However, because we only
measure size distribution, we still recommend storing at –20 °C for
long periods of time and limiting freeze-thaw cycles, particularly if
the measurement method requires long, undamaged DNA fragments.
Homogeneity: NIST sequenced multiple vials in an experiment designed
to assess homogeneity of the samples. No significant differences were
detected in terms of proportion of variant or copy number, except for
a few in regions known to be susceptible to systematic errors. These
results, along with the mixing of DNA before aliquoting, provide
confidence that no large differences in small variants or copy number
are likely to exist between different vials
I guess the extra QC steps and the fact that it's this sample that was used for GIAB variant calls would make RM9398 the gold standard if you were looking to validate your pipeline etc.
Oh wow, I thought I read that but apparently I glossed over the part where they wrote "DNA is currently available from this same cell line from Coriell with the number NA12878, but it may contain small differences in the DNA sequence due to different mutations occurring in different batches of the cells.".
Thanks for the response!