Terminal restriction fragment length measurement errors are affected mainly by fragment length, GÂ +Â C nucleotide content and secondary structure melting point

Several methods of molecular analysis of microbial diversity, including terminal restriction fragment length polymorphism (T-RFLP) analysis are based on measurement of the DNA fragment length. Significant variation between sequence-determined and measured length of restriction fragments (drift) has been observed, which can affect the efficiency of the identification of microorganisms in the analyzed communities. In the past, this variation has been attributed to varying fragment length and purine content. In this study, principal component analysis and multiple regression analysis were applied to find the contributions of those and several other fragment characteristics. We conclude that secondary structure melting point and G + C nucleotide content, besides the fragment length, contribute to the variation observed, whereas the contribution of purine content is less important. Incomplete denaturation of the sample at the start of electrophoretic separation of fragments has been excluded as a major cause of the variation observed. Our regression model explains the observed drift variation by approximately 56%, with standard deviation of the prediction equal to approximately 1.3 bp.