148. AptaMetrics: A Web-Based Aptamer Bioinformatics Platform

Abstract

RNA and DNA aptamer development as a therapeutic moiety and as a diagnostic tool is a rapidly growing field. The discovery process for aptamers involves iterative rounds of selection in a process termed SELEX, Systematic Evolution of Ligands by EXponentional enrichment. The modern SELEX process was revolutionized with the adoption of high-throughput sequencing (HTS), which is able to identify millions of aptamer sequences across multiple rounds of aptamer selection. The scale of aptamer HTS datasets has necessitated bioinformatics algorithms, such as clustering algorithms, to identify candidate aptamers for experimental validation. Herein we describe a web-based aptamer bioinformatics platform, AptaMetrics, which utilizes a collection of programs and algorithms to facilitate the analysis of aptamer HTS datasets.The AptaMetrics platform builds upon our previously described aptamer bioinformatics work (Thiel et. al. PLoSOne, 2012). Central to AptaMetrics is our clustering algorithm, ProcessSeqs, which we engineered to include several novel and distinct features: 1) clustering may be performed using either an “all-versus-all” or a “seed” sequence strategy; 2) clustering may be accomplished through either sequence similarity (edit distance) or structural similarity (tree distance); 3) the output of ProcessSeqs provides a visual and interactive representation of related aptamer networks; and 4) these networks of related aptamers may be analyzed using conventional network analysis tools. We also developed an aptamer structure prediction program, ProcessFold, that provides the predicted structure of millions of aptamer sequences utilizing a variety of proven oligonucleotide structure prediction algorithms (e.g., RNAfold, Mfold, RNAstructure). Integration of these two programs thereby allows users to create networks of structurally related aptamers based on the different structure prediction algorithms. We also created and incorporated an aptamer truncation algorithm, AptmR, into the AptaMetrics platform. The purpose of AptmR is to determine a predicted truncated aptamer sequence based on aptamer sequence and structure information provided by ProcessFold. AptmR can be applied to either a single aptamer sequence or to millions of aptamer sequences. The predicted truncations may then be clustered into networks of related aptamers using the ProcessSeqs clustering algorithm. Importantly, to allow the greatest versatility, each program and algorithm within AptaMetrics may be used independently or as a pipeline. The collection of programs and algorithms within AptaMetrics provide a significant degree of flexibility in analyzing aptamer HTS data while maintaining the ease of accessibility of a web-based environment.