Margin for Error in Paternity Testing

[aes001]

I'm not great when it comes to the sciences, but I joined this forum in hopes that those of you who know about genetics might be able to answer a rather important question for me.

 

I was adopted as a baby and told that my birth parents did not want me. The very abridged version of the story is that a few months ago I found out that my birth father hadn't actually known of my existence. Now that he does, we went ahead and got a paternity test done. The results came back excluding him as my father because four of fifteen alleles did not match. My gut reaction was that something had gone wrong at the lab; I look just like this man, it's uncanny. I began to research and found that false exclusions do happen--albeit rarely--due to things like contamination and human error. In all but one of the instances where we are unmatched it is only by one point...I know there wouldn't be three mutations but could contamination cause alleles to be off by a point or two? Either way, my feeling is that the test needs to be redone, just in case, but in the meantime I'm hoping that some of you may be able to educate me further on the topic of false exclusions within paternity testing. Thanks in advance for any and all help!

 

I've included the testing details below, as well as a copy of the test results:

 

--the company we went through is called easyDNA, though they outsource to GTL labs in New Mexico (which is affiliated with New Mexico State University)

--we used a 'peace of mind' test, so our samples were mailed in and did not follow the legal chain of custody

--this info may be irrelevant, but since seeing the results I've done a bit of digging and it seems that almost no one has ever heard of easyDNA or GTL

 

[CharonY]

Well, on the analytical side the error is very low (any half-way decent lab can perform it accurately). The only real issue is that samples could have been mixed up during sampling or processing (should also not happen).It wold be even more conclusive if the mother was also tested but it is more likely to (falsely) include a parent rather than exclude them. Four mismatches would require mutations in the respective sites and that amount is highly unlikely, as you already mentioned.

If the result is wrong it would really require mishandling of the samples. The lab is reputable (and is also offering services in the UK) so I doubt there would be a high likelihood of errors on their part. Obviously, a repetition should erase any remaining doubts (using the same set of test alleles).

[aes001]

Thanks for your help!