Methods for high fidelity production of long nucleic acid molecules

A method for removing or controlling errors in a plurality or pool of nucleic acid molecules, the method comprising the steps of:

• a) synthesizing said plurality or pool of nucleic acid molecules by the steps of:
• providing a first immobilized nucleic acid comprising a first 5′ region and a first 3′ region;
• providing a second immobilized nucleic acid comprising a second 5′ region and a second 3′ region, wherein said second 3′ region and said first 5′ region comprise identical nucleic acid sequences;
• hybridizing said first immobilized nucleic acid with an oligonucleotide under conditions promoting hybridization of said oligonucleotide to said first 3′ region,
• extending the hybridized oligonucleotide, and producing a first extension product comprising a 3′ region that is complementary to said first 5′ region;
• and
• hybridizing said second immobilized nucleic acid with said first extension product under conditions promoting hybridization of said 3′ region of said first extension product to said second 3′ region, extending the 3′ region of said first extension product, and producing a second extension product comprising a 3′ region that is complementary to said second 5′ region, wherein said second extension product comprises a sequence complementary to said first 3′ and 5′ regions and said second 3′ and 5′ regions, and the individual nucleic acid molecules of said second extension product comprise said plurality or pool of nucleic acid molecules;
• b) separating error-containing nucleic acid molecules of said plurality or pool of nucleic acid molecules from error-free nucleic acid molecules of said plurality or pool of nucleic acid molecules by contacting said plurality or pool of nucleic acid molecules with a component that actively selects for a sequence error under conditions permitting said error-containing nucleic acid molecules to bind with the component that actively selects for a sequence error, wherein the component that actively selects for a sequence error only specifically binds to said error-containing nucleic acid molecules;
• and
• c) removing or controlling errors in the plurality or pool of nucleic acid molecules by removing nucleic acid molecules that bind with the component that actively selects for a sequence error from the plurality or pool of nucleic acid molecules such that said error-containing nucleic acid molecules from said plurality or pool of nucleic acid molecules are separated from said error-free nucleic acid molecules.