Summary
Activity profiles (active vs inactive) of ancestrally reconstructed sequences were consistently less effectively distinguished using a large variety of metrics when compared to extant and ML-generated enzymes (1).
Figures
Figure 2A from (1)
See also
- Most ML quality metrics cannot effectively predict enzyme activity after controlling for similarity to native
- Ancestral sequence reconstruction outperforms generative ML methods on functional enzyme design
1.
Johnson SR, Fu X, Viknander S, Goldin C, Monaco S, Zelezniak A, et al. Computational scoring and experimental evaluation of enzymes generated by neural networks. Nature Biotechnology. 2024;43(3):396–405. Available from: https://doi.org/10.1038/s41587-024-02214-2