Why are repetitive sequences over-amplified by PCR during sequencing library preparation?

Question

[saarantras]

Like the title says. I've seen this mentioned as a problem before (for example, this page states "blacklisted (repeat) regions with ultra high signal will also be high in duplicates"). However I don't understand its chemical basis. Why wouldn't a repetitive sequence simply increase with 2^cycles during PCR, as non-repetitive sequences do? Why would PCR be biased to amplify repetitive sequences more than nonrepetitive sequences. Does this warning apply only to simple repeats, or also to genome-wide repeats (like retrotransposons)?

Answer 1

[Mensur Dlakic]

Not a bioinformatics question. As a general rule, I think you should be looking elsewhere for answers to these kinds of questions.

Individual non-repetitive sequence also amplify according to 2^cycles, but there are multiple copies of them. The final tally, at least theoretically, will be copy-number * 2^cycles. Realistically, after several cycles there will be so many copies of repetitive sequences that they will outcompete non-repeats for primers and polymerase, so the outcome may be ever greater in their favor than the theoretically expected ratio.