Data underlying the figures in the publication “Large-scale DNA-based phenotypic recording and deep learning enable highly accurate sequence-function mapping”, published in Nat Commun, 2020, 11, 3551. https://doi.org/10.1038/s41467-020-17222-4 Table of contents: 1. Table 1; Excel file with the source data for Figs. 1d, 2c–e, 3b, d, e, 4b–f, and 5a–c, e, f and Supplementary Figs. 1b, 3a, b, 4a, b, 5, 6b, 7a, b, 9a, b, 10, 11, 12a–c, 13, 14a, b, 15d, e, 16, 17a–d, 18, and 25. 2. uASPIre-master; .zip folder with NGS code generated and used in this study 3. SAPIENs-main; .zip folder with the ML code generated and used in this study Plasmids pASPIre1, pASPIre2, and pASPIre3 are available from Addgene under identification numbers ID154842, ID154843, and ID154843, respectively. The raw NGS data for the three biological replicate experiments for the large RBS library are provided in the NCBI SRA under accession codes SAMN15215026, SAMN15215027, and SAMN15215027. The processed data sets and codes are also available under: github.com/JeschekLab/uASPIre (NGS data processing) and github.com/BorgwardtLab/SAPIENs (code for ML).