Table 1.
Performance overview of single‐site saturation mutagenesis methods (e. g. used for deep mutational scanning).
|
Method |
Protein (# of mutated codons) |
% Coverage[a] |
% 1 NSM[b] |
% 0 NSM (wild‐type)[c] |
% >1 NSM (off‐target)[d] |
Ref. |
|---|---|---|---|---|---|---|
|
in vitro | ||||||
|
Nicking mutagenesis (NM) |
E. coli AmiE (70)[e] |
100 |
36.4 |
52.7 |
10.8 |
[36] |
|
A. thaliana PYR1 (86)[e] |
100 |
26.5 |
59.4 |
13.6 |
[36] |
|
|
Anti‐influenza human antibody variable heavy gene UCA9 (99)[e] |
97.4 |
14.1 |
71.3 |
15.0 |
[36] |
|
|
Phage ΦX174 F capsid protein (421) |
100 |
41.8 |
28.6 |
29.5 |
[50] |
|
|
Phage ΦX174 G spike protein (172) |
99.9 |
49.3 |
29.3 |
21.4 |
[50] |
|
|
PALS |
S. cerevisiae Gal‐DBD (64) |
99.9 |
47 |
24 |
21 |
[37] |
|
Human p53 (393) |
93.4 |
33 |
30 |
35 |
[37] |
|
|
| ||||||
|
in vivo | ||||||
|
Plasmid recombineering (PR)[f] |
A. thaliana‐derived iLOV (110) |
99.8 |
28.8 |
60.7 |
10.5 |
[41] |
|
CREATE (genomic mutagenesis) |
E. coli GalK (1) |
100 |
56.8[g] |
22.4[g] |
n/a |
[42] |
|
S. cerevisiae ADE2 (2) |
100 |
95 |
5 |
n/a |
[42] |
|
|
E. coli lysine metabolism, 19 genes, 16,300 designed edits[h] |
22.7 to 61.6 |
n/a |
n/a |
n/a |
[54] |
|
n/a: not available; [a] Number of actually observed mutations per 100 designed mutations. Note: differences in sequencing depth used for quality control in the different studies influences the number of observed mutations, thus the apparent coverage. [b] Percent of mutants that carry exactly one desired non‐synonymous mutation (NSM). [c] Percent of mutants that do not carry any NSM, thus being wild‐type (non‐edited) variants. Note that CREATE is the only method that counter‐selects for wild‐type via CRISPR‐Cas9 mediated double‐strand breaks. [d] Percent of mutants that carry more than one NSM, e. g. off‐target mutations. [e] The average of two reported independent runs of NM is given (Table 1 in Ref. [36]). [f] Here, hand‐mixed pools of column‐synthesized oligonucleotides were used. It was shown later that PR also works with array synthesized oligo pools. [53] [g] Calculated based on 80 % of clones being edited (based on colorimetric screen) and 71 % of those 80 % being correctly edited (based on sequencing, 0.8×0.71); and 20 % of clones being wild‐type (based on colorimetric screen) plus 3 % of the 80 % phenotypically edited clones (0.8×0.03) being still unedited at the programmed locus. [h] CREATE was developed and applied for multiplexed pathway mutagenesis. The percent coverage refers to the observed coverage range of five test loci that were analyzed in‐depth (Table 1 in Ref. [54]).