DeepHeal: An Interpretable Framework for Phenotype-Based Antineoplastic Drug Discovery Using Drug-Perturbation Proteomics
4 affiliations
- 1Institute of Digital Materia Medica & Structural Pharmacology, College of Pharmacy
- 2Fujian University of Traditional Chinese Medicine
- 3Wuya College of Innovation
- 4Shenyang Pharmaceutical University
Abstract
Drug-perturbation proteomics provides a direct molecular readout of compound-induced cellular states, but its high dimensionality and limited interpretability hinder its use for mechanism-oriented drug discovery. Here, we present DeepHeal, an interpretable consensus representation framework that transforms drug-induced proteomic responses into biologically readable and experimentally actionable antineoplastic response programs. Using an HCT116 perturbation atlas comprising 875 compounds and 9,960 protein features, DeepHeal integrates ensemble variational autoencoders, supervised contrastive learning, consensus latent-variable aggregation, and protein-level attribution. A label-independent Average branch captures recurrent proteomic variation, whereas a complementary label-aware SupCon branch concentrates known antineoplastic structure. The resulting latent programs are mapped back to proteins, pathways, and interaction networks, revealing shared cell-cycle, DNA-replication, stress-response, metabolic, proteostasis, apoptosis-related programs, together with mechanism-dependent response heterogeneity. DeepHeal prioritized 19 initially unlabeled compounds for experimental follow-up, of which seven inhibited HCT116 cells with IC₅₀ values below 10 μM, including two compounds with nanomolar activity. Without retraining, DeepHeal representations further recovered antineoplastic response states across 566 external proteomic perturbation profiles spanning multiple cellular contexts. Together, these findings establish DeepHeal as, to our knowledge, the first proteomics-first interpretable consensus representation framework that transforms large-scale drug-perturbation proteomes into protein- and pathway-resolved response programs for cross-context anticancer candidate prioritization and experimental validation.
Related
- A transferable perturbation proteomics framework for phenotype- 1 based annotation and antiproliferative potency scoring of antineoplastic agents
- Foundation model enables interpretable open and error-tolerant searching for mass spectrometry-based proteomics
- General many-body perturbation framework for moiré systems
- High-Multiplex Proteomics: Analytical Trade-Offs and a Practical Selection Framework for Emerging Platforms.
- Validating Spectral Quality Measures as Proxies for Task Performance: A Controlled-Perturbation Framework for Raman Spectroscopy
BibTeX
@misc{feng2026deepheal,
title = {{DeepHeal: An Interpretable Framework for Phenotype-Based Antineoplastic Drug Discovery Using Drug-Perturbation Proteomics}},
author = {Disheng Feng and Jing Liang and Kexin Zhang and Xu Cai and Yueying Yang and Meijing Wang and Hua Li and Lixia Chen},
howpublished = {ChemRxiv},
year = {2026},
month = sep,
doi = {10.26434/chemrxiv.15000952/v2},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15000952/v2},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - DeepHeal: An Interpretable Framework for Phenotype-Based Antineoplastic Drug Discovery Using Drug-Perturbation Proteomics
AU - Disheng Feng
AU - Jing Liang
AU - Kexin Zhang
AU - Xu Cai
AU - Yueying Yang
AU - Meijing Wang
AU - Hua Li
AU - Lixia Chen
PY - 2026
DA - 2026/09/01
PB - ChemRxiv
DO - 10.26434/chemrxiv.15000952/v2
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15000952/v2
AB - Drug-perturbation proteomics provides a direct molecular readout of compound-induced cellular states, but its high dimensionality and limited interpretability hinder its use for mechanism-oriented drug discovery. Here, we present DeepHeal, an interpretable consensus representation framework that transforms drug-induced proteomic responses into biologically readable and experimentally actionable antineoplastic response programs. Using an HCT116 perturbation atlas comprising 875 compounds and 9,960 protein features, DeepHeal integrates ensemble variational autoencoders, supervised contrastive learning, consensus latent-variable aggregation, and protein-level attribution. A label-independent Average branch captures recurrent proteomic variation, whereas a complementary label-aware SupCon branch concentrates known antineoplastic structure. The resulting latent programs are mapped back to proteins, pathways, and interaction networks, revealing shared cell-cycle, DNA-replication, stress-response, metabolic, proteostasis, apoptosis-related programs, together with mechanism-dependent response heterogeneity. DeepHeal prioritized 19 initially unlabeled compounds for experimental follow-up, of which seven inhibited HCT116 cells with IC₅₀ values below 10 μM, including two compounds with nanomolar activity. Without retraining, DeepHeal representations further recovered antineoplastic response states across 566 external proteomic perturbation profiles spanning multiple cellular contexts. Together, these findings establish DeepHeal as, to our knowledge, the first proteomics-first interpretable consensus representation framework that transforms large-scale drug-perturbation proteomes into protein- and pathway-resolved response programs for cross-context anticancer candidate prioritization and experimental validation.
ER -