Summary

Concatenating features from ESM-1b and GEARNet was found to be less effective than either cross-attention, ESM-to-GearNet, or even just ESM-1b embeddings alone (1). Meanwhile, (2) found that concatenation was outperformed by the use of ResNet as an autoencoder bottleneck.

Figures

MethodF_maxAUPR
ESM-1b0.8640.889
ESM-GearNet
- w/ parallel fusion0.7330.759
- w/ series fusion0.8830.890
- w/ cross fusion0.8800.893

Ref (1)

1.
Zhang Z, Wang C, Xu M, Chenthamarakshan V, Lozano A, Das P, et al. A Systematic Study of Joint Representation Learning on Protein Sequences and Structures. 2023; Available from: https://arxiv.org/abs/2303.06275
2.
Detlefsen NS, Hauberg S, Boomsma W. Learning meaningful representations of protein sequences. Nature Communications. 2022;13(1). Available from: https://doi.org/10.1038/s41467-022-29443-w