Oligonucleotide-library design

The custom oligonucleotide pool containing pairs of sgRNA and corresponding target sequences was purchased from Twist Bioscience. The library includes 23,123 random DNA sequences and 5,171 disease loci theoretically targetable using base editors. Designed oligonucleotides include the following elements: The G/20N spacer and SpCas9 gRNA scaffold, a six-nucleotide randomized barcode, the corresponding target locus containing the PAM, and a second six-nucleotide randomized barcode (Supplementary Note ⁴). Randomized DNA sequences of 20 bp length and 1:1:1:1 proportion of each nucleotide were generated using a custom Python script to form a random sequence library. The disease loci were selected from the NCBI ClinVar¹⁸ database (accessed in May 2019) using the following criteria: (a) all disease-associated SNPs were accessed and restricted to pathogenic and monogenic filters (b) SNPs were further restricted to the possible base conversions targetable by ABEs (A-to-G) and CBEs (C-to-G). (c) Genomic region flanking the SNP genomic coordinates were extracted from UCSC server (http://genome.ucsc.edu/). (d) The sequences were then scanned presence of an NGG PAM 8–18 bases away from the target base. Only SNPs passing these filtering criteria were included in the study and were then appended to the list of aforementioned random sequences to form the final library.