ESPeR-seq: Extremely Sensitive and Pure, End-to-end, RNA-seq library preparation

Abstract

PCR-based RNA sequencing methods generate phantom unique molecular identifiers (UMIs) when residual UMI-bearing oligonucleotides reprime during amplification, inflating molecule counts and corrupting relative abundances. They also incur non-specific amplification that limits sensitivity, while synthetic poly-T tracts disrupt sequencing and obscure transcript 3' ends. Here we present Omega-seq, which restricts UMI incorporation to reverse transcription using uracil-containing, USER-excisable template-switching oligonucleotides and a uracil-intolerant polymerase. Omega-seq produces near-background-free libraries, distinguishes 10 fg of input RNA from negative controls, and eliminates bead cleanup. An Omega-shaped dT primer prevents poly-T-induced dephasing, enabling stranded, nucleotide-resolution mapping of transcript ends. Dilution series, ERCC spike-ins, and a rarefaction metric show that Omega-seq has the lowest UMI inflation among methods tested. In individual Drosophila neuroblasts, phantom-UMI inflation remains minimal; pooled neuroblast and neuron coverage enables de novo discovery of novel multi-exon genes, 3' UTR extensions, and candidate enhancer RNAs.