Table 2 Top-5% scores on the ranking task. MoMS-Net demonstrates outperform in FT-CID dataset, and comparable performance to MassFormer on FT-HCD dataset, respectively.

From: Mass spectra prediction with structural motif-based graph neural networks

 

FT-CID

FT-HCD

CNN

\(0.802\pm 0.008\)

\(0.778\pm 0.004 \)

MassFormer

\(\mathbf {0.850}\pm 0.016\)

\(0.830\pm 0.007 \)

WLN

\(0.736\pm 0.011\)

\(0.812\pm 0.008 \)

GCN

\(0.728\pm 0.0.016\)

\(0.802\pm 0.008 \)

MoMS-Net (Ours)

\({0.824\pm 0.002}\)

\(\mathbf {0.840\pm 0.010 }\)

  1. Every result is repeated five times with different random seeds. Significant values are in bold.
Back to article page