Table 1 Summary of simulated datasets. For each dataset, we list the library name (Library), sequence length in nucleotides (L), number of unique library sequences (\(M'\)), epistasis hyperparameter used for fitness simulation (T), read type (short, long, or hybrid), \(\%\) of the sequence of interest covered by individual reads (Cover), and number of pre-selection and post-selection reads (\(N^\text {pre}\) and \(N^\text {post}\)), which were always equal. We simulate \(4.6 \times 10^7\) short reads to match the experimental data from Zhu et al. [20], and up to \(4.6 \times 10^5\) long reads to be within the current throughput of PacBio’s technologies [34, 35]. Each dataset is described in more detail in the Methods
From: MBE: model-based enrichment estimation and prediction for differential sequencing data
Library | L | M′ | T | Read type | Cover | \(N^\text {pre}=N^\text {post}\) |
|---|---|---|---|---|---|---|
21-mer insertion | 21 | \(8.5 \times 10^6\) | 140 | Short | 100 | \(4.6 \times 10^7\) |
150-mer insertion | 150 | \(8.5 \times 10^6\) | 1000 | Short | 100 | \(4.6 \times 10^7\) |
300-mer insertion | 300 | \(8.5 \times 10^6\) | 2000 | Short | 100 | \(4.6 \times 10^7\) |
avGFP mutagenesis | 714 | \(2.5 \times 10^7\) | 4760 | Long | 100 | \(4.6 \times 10^5\) |
avGFP mutagenesis | 714 | \(2.5 \times 10^7\) | 4760 | Short | 42 | \(4.6 \times 10^7\) |
AAV recombination | 2253 | \(2.6 \times 10^{7}\) | 15,020 | Long | 100 | \(4.6 \times 10^5\) |
AAV recombination | 2253 | \(2.6 \times 10^{7}\) | 15,020 | Long | 100 | \(4.6 \times 10^4\) |
AAV recombination | 2253 | \(2.6 \times 10^{7}\) | 15,020 | Long | 100 | \(4.6 \times 10^3\) |
AAV recombination | 2253 | \(2.6 \times 10^{7}\) | 15,020 | Short | 13 | \(4.6 \times 10^7\) |
AAV recombination | 2253 | \(2.6 \times 10^{7}\) | 15,020 | Hybrid | 100 long + 13 short | \(4.6 \times 10^3\) long + \(4.5 \times 10^7\) short |