If you’re thinking “didn’t we do this decades ago?” the answer is “not really.” They were sampled and statistical inferences used to map them.
That’s not quite the issue. This sequence was also created by sampling and assembling the raw sequences with statistical algorithms. You can’t really get away from that. What this version used were reads that were much longer than the ones used twenty years ago and so the the assembly should be more accurate and include the very repetitive telomere sequences that are notoriously hard to assemble with short reads.
