If we *had* to order the amino acids by similarity along one dimension, how would we do it? I’ve come up with a starting point that looks… at least like a good starting point. Using the RNA codon table I used the second nucleotide as most significant, the first nucleotide after that and the least significant being the third. Serine was a bit of a wild card, but this is the order I derived:

1. G
2. R
3. W
4. C
5. E
6. D
7. K
8. N
9. Q
10. H
11. Y
12. A
13. T
14. P
15. S
16. V
17. M
18. I
19. L
20. F