Single-mutation scan
Every possible single mutation scored by the model: one column per residue, one row per amino acid. Blue means the model predicts a better score than the wild type, red a worse one.
avgfp · Sequence CNN
worse than WT
better than WT
Hover a cell for its value. Diagonal hatching marks the wild-type amino acid at that position, which is not a mutation. Raw values (JSON)
Highest-scoring mutations
| # | Mutation | Predicted score | vs wild type |
|---|---|---|---|
| 1 | Y37R |
4.168 | +0.449 |
| 2 | N103W |
4.064 | +0.345 |
| 3 | S203T |
4.038 | +0.319 |
| 4 | N103F |
4.020 | +0.300 |
| 5 | N103Y |
4.018 | +0.299 |
| 6 | N103Q |
4.013 | +0.294 |
| 7 | A152P |
4.004 | +0.284 |
| 8 | E15R |
4.000 | +0.280 |
| 9 | H229P |
3.998 | +0.279 |
| 10 | V161P |
3.996 | +0.277 |
| 11 | G189N |
3.994 | +0.275 |
| 12 | L13R |
3.987 | +0.267 |
| 13 | N103L |
3.986 | +0.266 |
| 14 | M231E |
3.985 | +0.266 |
| 15 | V161A |
3.972 | +0.253 |
| 16 | N103M |
3.970 | +0.250 |
| 17 | N103K |
3.965 | +0.245 |
| 18 | G230P |
3.963 | +0.244 |
| 19 | Q202Y |
3.958 | +0.239 |
| 20 | V161G |
3.958 | +0.238 |
| 21 | Y37N |
3.953 | +0.234 |
| 22 | H229N |
3.952 | +0.233 |
| 23 | A108S |
3.952 | +0.233 |
| 24 | G8S |
3.951 | +0.232 |
| 25 | G189R |
3.944 | +0.225 |
| 26 | Q155P |
3.943 | +0.224 |
| 27 | E170L |
3.940 | +0.221 |
| 28 | T36H |
3.939 | +0.220 |
| 29 | N103H |
3.939 | +0.220 |
| 30 | T201I |
3.938 | +0.219 |
These are predictions, not measurements. Single mutants are a small part of the training data for most of these datasets, so treat the list as candidates to test.