Critical Mass: Structure-Disjoint Splits Drift With the Libraries They Are Built On

Key points

• For predicting structures from mass spectra, the field has converged on single-linkage clustering under the myopic Maximum Common Edge Subgraph (MCES) distance at a threshold of 10.1
• We propose an optimized version of this criterion, allowing for computationally feasible application on the public libraries: Admissible bounds and a union-find skip reduce $2.6 \times 10^{10}$ molecular pairs to only $1.15 \times 10^{7}$ integer programs, a $2{,}259\times$ saving, while provably producing the lossless partition.1
• We find that at this scale the graph percolates.1
• Large library benchmarks using this splitting rule naively therefore conflate structural novelty with extrapolation to larger, more complex chemistry.1

Sources (1)

• [1]Critical Mass: Structure-Disjoint Splits Drift With the Libraries They Are Built On
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 08:20 PM

  Molecular machine learning evaluates out-of-distribution performance with structure disjoint splits, a construction rule fixed once at the scale its library happened to have: threshold a pairwise distance, then assign whole connected components to folds.

  For predicting structures from mass spectra, the field has converged on single-linkage clustering under the myopic Maximum Common Edge Subgraph (MCES) distance at a threshold of 10.

Extractive summary: sentences quoted from the sources.