To test a variant of interest, both the reference and the alternative allele were designed into 200nt oligos synthesized by Twist Biosciences. Each oligo contains a 164nt flanking sequence centered around the variant and adaptor sequences for cloning. The oligos were then cloned into the 3’ UTR region of the eGFP gene, whose expression is driven by the CAG promoter, in the plasmid reporters. We introduced 3ug plasmid libraries into 15M HEK293/HeLa cells via cell electroporation for each biological replicate, with a total number of three replicates. Total mRNA was isolated from the cells 24h after electroporation, reverse-transcribed into cDNA, and made into sequencing libraries through a stepwise PCR. Specifically, a unique molecular identifier (UMI) was added to each cDNA transcript in the first round of PCR. The PCR products were further amplified to add Illumina sequencing adaptors in the second round of PCR. A similar stepwise PCR protocol was also applied to the plasmid DNA to generate DNA sequencing libraries. Both DNA and RNA sequencing libraries were pooled together to be sequenced on Illumina HiSeq 3000 PE150 or NovaSeq 6000 PE150 with 15% PhiX spike-in. For data analysis, we extracted UMIs from each read to remove PCR duplicates. The remaining reads were aligned to the designed reference sequences and UMIs were counted using UMItools¹. UMI counts were normalized before calling functional variants using MPRAnalyze².

Reference

1. Smith, T., Heger, A. & Sudbery, I. UMI-tools: Modeling sequencing errors in Unique Molecular Identifiers to improve quantification accuracy. Genome Res. 27, 491–499 (2017).

2. Ashuach, T. et al. MPRAnalyze: Statistical framework for massively parallel reporter assays. Genome Biol. 20, 1–17 (2019).