>>11685879DNA polymerase will on average make a mistake every 1-2kb or so (usually just single-nucleotide change). there are a number of mechanisms for repairing this, especially when considering that a basepair change like that will usually result in a distortion of the strand due to non-Watson-Crick base pairing, which is recognized by mismatch repair (MMR) machinery. However every so often one goes unchanged, and in this case the repair machinery might not 'get' which one was mutated, in which case it'll do one of several things: 1. randomly switch it for a T (especially if the complementary base is an A) 2. base pair excision, basically one base pair is snipped and they try to match it to the other one, sometimes it 'guesses' the right one correctly, sometimes not 3. chew back a few bases and try again, i.e. NHEJ 4. if it's post s-phase, it might use homology directed repair using the recently synthesized sister chromatid as template 5. fucking flip out, inhibit MDM2, unleash p53, initiate apoptosis cascade (this is unlikely for just a single base-pair mistake)
this isn't comprehensive but basically the machinery isn't perfect and mistakes happen. is there a pattern to the mistakes? yeah i guess if you count fidelity statistics to be a 'pattern' but really if a motor misfires every 1000-2000 hours of run-time, does that count as a 'pattern'? kinda, yeah, but also not necessarily a distinct and measurable pattern, just an average rate of mistakes (for instance what if one misfire happens 100 hours in, or 4000 hours go by without one - outliers against a statistical trend, but does that mean they "invalidate the pattern"? not really). if a clerk makes a mistake and misses a number every 1000-2000 lines or so, same concept. is it a pattern or is it just the limitations of one's equipment?
