Abstract A33: Optimizing NGS-based shRNA screening experiments and data analysis

Pooled small hairpin RNA (shRNA) screening emerges as a powerful tool for loss-of-function studies in mammalian cells and searching for novel or synthetic lethality-based therapeutic targets of human tumors. Next-generation sequencing (NGS) is becoming a standard technology to read out hairpin representations with high resolution and allowance of massive multiplexing. However, the extraction of robust biological signals from high-throughput NGS-based shRNA screening (shSeq) data remains challenging due to low knock-down efficacy and off-target effects of shRNAs, sequencing errors, artificial mutations, etc. In this presentation, based on our in vitro shSeq screens in human cancer cells by Illumia MiSeq, we report a discovery of an unexpected recurrent distribution pattern of identified hairpin counts across number of allowed mismatches during read deconvolution. By exhaustive analysis of within and between-group noise, we develop a procedure to optimize both experimental and analytical parameters for shSeq screens and suggest a set of default parameters that perform well consistently. We observe that over 10% of recovered hairpin reads have at least one mismatch from the reference sequence, which cannot be explained by sequencing error only. We also notice that there are more mismatched reads at later time point than at initial one. And computational analysis of such dynamic process reflects higher evolutionary mutation rate (>10^-5) in cancer cells than common sense (~10^-9). Artificial mutations that might cause a high percentage of mismatches are further explored and interesting mutation patterns are identified. Citation Format: Jiyang Yu, Ying Gao, Wenyang Zhong, Lei Chen, Donald Apanovitch, Max Follettie, Paul Rejto, Kim Arndt. Optimizing NGS-based shRNA screening experiments and data analysis. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Synthetic Lethal Approaches to Cancer Vulnerabilities; May 17-20, 2013; Bellevue, WA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(5 Suppl):Abstract nr A33.