A novel strategy for retention prediction of nucleic acids with their sequence information in ion-pair reversed phase liquid chromatography.
Journal:
Talanta
Published Date:
Apr 10, 2018
Abstract
In this work, retention behaviors of oligonucleotides and double-stranded deoxyribonucleic acids (dsDNAs) have been investigated in ion-pair reversed-phase liquid chromatography (IP-RPLC). We demonstrated that classic linear solvent strength (LSS) model is applicable for describing isocratic retention of oligonucleotides and dsDNAs, which indicated that nucleic acids share the similar retention mechanism as other common small molecules in IP-RPLC. The separation of nucleic acids in IP-RPLC is driven by both hydrophobic and electrostatic interactions. We defined the parameter lnk/S obtained from LSS model in IP-RPLC as chromatographic hydrophobic and electrostatic interaction index (CHEI). CHEI of a nucleic acid has been revealed to correlate well with its gradient retention time. Notably, we proposed a strategy for retention prediction based on CHEI and base sequence information of nucleic acids. Corresponding to base locus, each base sequence is converted to a featured locus vector consisting of zeros and ones. CHEI prediction models were established by support vector regression (SVR) algorithm with locus vectors. Predicted CHEI values have been applied to predict retention times under desired gradient elution runs. This protocol is easy to grasp and worth pursuing further development for more precise retention prediction performance of nucleic acids in IP-RPLC.