TY - JOUR T1 - RASLseqTools: open-source methods for designing and analyzing <span class="underline">R</span>NA-mediated oligonucleotide <span class="underline">A</span>nnealing, <span class="underline">S</span>election, and, <span class="underline">L</span>igation <span class="underline">seq</span>uencing (RASL-seq) experiments JF - bioRxiv DO - 10.1101/036061 SP - 036061 AU - Erick R. Scott AU - H. Benjamin Larman AU - Ali Torkamani AU - Nicholas J. Schork AU - Nathan Wineinger AU - Max Nanis AU - Ryan Thompson AU - Reza B. Beheshti Zavareh AU - Luke L. Lairson AU - Peter G. Schultz AU - Andrew I. Su Y1 - 2016/01/01 UR - http://biorxiv.org/content/early/2016/01/07/036061.abstract N2 - RNA-mediated oligonucleotide Annealing, Selection, and Ligation (RASL-seq) is a method to measure the expression of hundreds of genes in thousands of samples for a fraction of the cost of competing methods. However, enzymatic inefficiencies of the original protocol and the lack of open source software to design and analyze RASL-seq experiments have limited its widespread adoption. We recently reported an Rnl2-based RASL-seq protocol (RRASL-seq) that offers improved ligation efficiency and a probe decoy strategy to optimize sequencing usage. Here, we describe an open source software package, RASLseqTools, that provides computational methods to design and analyze RASL-seq experiments. Furthermore, using data from a large RRASL-seq experiment, we demonstrate how normalization methods can be used for characterizing and correcting experimental, sequencing, and alignment error. We provide evidence that the three principal predictors of RRASL-seq reproducibility are barcode/probe sequence dissimilarity, sequencing read depth, and normalization strategy. Using dozens of technical and biological replicates across multiple 384-well plates, we find simple normalization strategies yield similar results to more statistically complex methods. ER -