Metraipolis › archive › dc_44aaa973e8cd

Feature Space Selection and Heterogeneous Effect Estimation for Blood-Brain Barrier Permeability: A Random Forest to the Generalized Random Forest Pipeline — arXiv:2609.29076

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.29076
License (as archived)
CC0-1.0
Archived
2026-10-08T05:26:28.854Z
SHA-256 of the archived text
57dd9a2c5d699ff94231587673c33c96137d778b169e83077197a67074cee06b
Stable URL of this page
https://metraipolis.com/documento/dc_44aaa973e8cd

How to cite

Feature Space Selection and Heterogeneous Effect Estimation for Blood-Brain Barrier Permeability: A Random Forest to the Generalized Random Forest Pipeline — arXiv:2609.29076 — arXiv (en), https://arxiv.org/abs/2609.29076 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_44aaa973e8cd (Metraipolis archive dc_44aaa973e8cd, sha256 57dd9a2c5d699ff94231587673c33c96137d778b169e83077197a67074cee06b)
BibTeX
@misc{dc44aaa973e8cd,
  title = {Feature Space Selection and Heterogeneous Effect Estimation for Blood-Brain Barrier Permeability: A Random Forest to the Generalized Random Forest Pipeline — arXiv:2609.29076},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.29076},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_44aaa973e8cd; sha256 57dd9a2c5d699ff94231587673c33c96137d778b169e83077197a67074cee06b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.29076. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Predicting blood-brain barrier (BBB) permeability is critical for central nervous system drug discovery. Using the MoleculeNet BBBP dataset (n = 2039), this study systematically ablates molecular feature spaces to isolate featurisation from model architecture. We evaluate three feature families (Morgan fingerprints, RDKit physicochemical descriptors, SMILES bigrams) across four learning algorithms. Results demonstrate that predictive performance depends jointly on feature representation and algorithm. Dynamic Random Forest using combined features achieved the highest mean AUC (0.970, 95% CI: 0.963-0.977). Second, this optimal representation enables exploratory estimation of heterogeneous associations between molecular structure and BBB permeability using Generalized Random Forests. Constructing a pseudo-treatment from a LogP median split, we applied double/debiased machine learning to account for confounding. Orthogonalization substantially attenuates the heterogeneity detected by naive causal forests; no conditional effects remained significant after false discovery rate correction (smallest adjusted p = 0.082). Furthermore, orthogonalized feature importance shifted toward residual structural information in SMILES bigrams. Ultimately, once observed confounding is properly accounted for, evidence that LogP-BBB associations vary systematically across chemical space is insufficient. This underscores that feature representation and model architecture are coupled design choices, and that unorthogonalized causal forests risk overstating genuine treatment effect heterogeneity.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-25 in cs.LG).
Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.29076 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T04:50:39.587Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).