Proof-reading
Proof-reading is a process that involves several proof-reading enzymes to detects and correct errors found in DNA replication, transcription, and translation[1].
During a semi-conservative DNA replication, a new complementary strand of DNA strand is being formed from a DNA template strand by complementary base-pairing. However, apart from Watson-Crick base pairs (A bind to T and G bind to C), there are several types of non-Watson-Crick base pairs. These non-Watson-Crick base pairs causes different nucleotides being added to the newly synthsised strand, resulting in a different DNA strand being formed. A single change of nucleotides in the sequence may cause an enormous effect to the organism such as malfunction of the proteins made. The condition is worsen if this change became permanent[2].
With the help of proof-reading process, this mechanism has high accuracy as only around 1 mistake occurs for every 109 (1 Giga) nucleotides paired.
In the proof-reading (error correcting) process, DNA polymerases remove incorrectly paired nucleotides with the help of 3'-to5' proofreading exonuclease.