my midas cow is clean.
Right DL on not ever being able to prove. there isn't much statistical difference between a 1 in 10 billion vs a 1 in a 100 billion chance that OJ wasn't the killer, but it was a mistake in the trial that my old stats professor made that was enough to make the jury "doubt" the dna evidence.
In our science seminars, we always say, you can never prove anything, you can only fail to disprove. that's why it's always important to maintain the idendity of the samples as they move through the process, and pedrigrees, frankly are not a good repositor. I find it just as important to maintain sample integrity as expermental quality. so does fornensics, which is why that would be a good job to have if you can put up with the court time and not having to deal with people like nifong. lots of sample id's have caused lots of problems in sequencing the human geneome, particularly in the early days when samples moved between centers and contamination for other genomes got into the mix. interesting side bar here is that even thought the genome is finished, there are lots of undiscovered SNP's yet to be discovered because it is not possible to sequence the same haploid representative on all the genome. even between the two haploids there will be snps between each other and during the discovery phase, some will be missed depending on which haploid was used. This is one reason Maines and shorthorns were so lucky that only were PHA and TH single genes, but that probably both haploids were used to help discover the gene becuase the "cow" wasn't sequenced yet and a mapping process was used, if not, then luckier still. barrel racer could elaborate on that some day when the PHA paper can come out. that's going to be big news.
