Skip to content
DraconiticResearch
Search Sources API
186,086 studies31 sourcesUpdated 44 min ago
Peer-reviewedCC BY-NC-NDJournal of chemical information and modeling

Deep3DCCS: Geometry-Driven Collision Cross Section Prediction from Multi-View Molecular Projection Tensors.

  • Pattipong Wisanpitayakorn1,2,3
  • Vivek Bhakta Mathema1,2,3
  • Kwanjeera Wanichthanarak1,2,3
  • Thanyanatthamon Keawhawong2,3
  • Narumol Jariyasopit1,2,3
  • Pairash Saiviroonporn4
  • Trongtum Tongdee4
  • Yongyut Sirivatanauksorn2,3
  • Sakda Khoomrung1,2,3,5,6
6 affiliations
  1. 1Siriraj Center of Research Excellence in Metabolomics and Systems Biology (SiCORE-MSB), Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok10700, Thailand.
  2. 2Siriraj Metabolomics and Phenomics Center, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok10700, Thailand.
  3. 3Thailand Metabolomics Society, Bangkok10700, Thailand.
  4. 4Department of Radiology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok10700, Thailand.
  5. 5Department of Biochemistry, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok10700, Thailand.
  6. 6Center of Excellence for Innovation in Chemistry (PERCH-CIC), Faculty of Science Mahidol University, Bangkok10400, Thailand.
Read full textPDFDOICite

Abstract

Collision cross section (CCS) measurements from ion mobility-mass spectrometry improve confidence in molecular identification in metabolomics and lipidomics and reduce false-positive annotations in untargeted analyses. However, experimental CCS coverage remains incomplete, and accurate computational prediction remains needed. Here, we present Deep3DCCS, a geometry-driven deep learning framework that predicts CCS from stacked multi-angle molecular projection tensors using a three-dimensional convolutional neural network operating across the projection-view and two spatial image dimensions. Three-dimensional molecular structures generated from SMILES strings were rotated around the x, y, and z-axes, converted into binary two-dimensional projections, and assembled into standardized multi-view tensors for CCS regression. Deep3DCCS was developed and evaluated using curated AllCCS and METLIN lipid CCS datasets containing 3948 molecule-adduct entries across [M - H]-, [M + H]+, and [M + Na]+ ions, and input optimization identified 32 × 32 pixel projections with five rotations per axis as a practical balance between prediction accuracy, model stability, and computational feasibility. In independent true-validation experiments, AllCCS-trained models achieved MRPE values of 3.27-3.79% on AllCCS validation data and 2.93-3.72% on METLIN validation data, whereas METLIN-trained models transferred less effectively to the broader AllCCS chemical space. Overall, these results support multi-view molecular projection tensors as a viable geometry-driven representation for CCS prediction and demonstrate that substantial CCS-relevant information can be learned directly from three-dimensional molecular morphology using a 3DCNN architecture, with low single-digit prediction errors across independent and cross-dataset validation experiments.

Related

  1. Modelling Energy Transfer Efficiency in Collision Induced Dissociation and Unfolding Using Molecular Dynamics Simulations
    Austin W. Green, James S. Prell
    Journal of the American Society for Mass SpectrometryPeer-reviewedLicence not statedFull textPhysical chemistryPMC
  2. Translation-rotation internal coordinates with splicing (TRICS): A new approach to molecular geometry interpolation
    Heejune Park, Chenchen Song, Lee-Ping Wang
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv
  3. Geometry‐Based Neural‐Network Prediction of Electron Localization Function Topology in Dense Hydrogen
    Xiaoyu Wang, Miriam Marqués, Sergio Gómez +3
    Chemistry - A European JournalPeer-reviewedCC BY 4.0Full text1 citedTheoretical & computational chemistryHALPMC
  4. Computational Evaluation of Carbazole–Chalcone Urea and Thiourea Derivatives Against EGFR: Molecular Docking, Pharmacokinetic Profiling, Target Prediction, and Molecular Dynamics Simulation
    Arleta Rifati-Nixha, Mustafa Arslan, Miribane Dërmaku-Sopjani +1
    Molecules (Basel, Switzerland)Peer-reviewedCC BYFull textGeneral chemistryEurope PMC
  5. A Machine Learning Approach to logBB Prediction Combining Molecular Docking and Quantum-Chemical Descriptors
    Lilian Rouille
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv

Cite

BibTeXRIS
BibTeX
@article{wisanpitayakorn2026deep3dccs,
  title = {{Deep3DCCS: Geometry-Driven Collision Cross Section Prediction from Multi-View Molecular Projection Tensors.}},
  author = {Pattipong Wisanpitayakorn and Vivek Bhakta Mathema and Kwanjeera Wanichthanarak and Thanyanatthamon Keawhawong and Narumol Jariyasopit and Pairash Saiviroonporn and Trongtum Tongdee and Yongyut Sirivatanauksorn and Sakda Khoomrung},
  journal = {Journal of chemical information and modeling},
  year = {2026},
  month = sep,
  doi = {10.1021/acs.jcim.6c00371},
  url = {https://europepmc.org/article/MED/42906041},
  note = {CC BY-NC-ND},
}
RIS
TY  - JOUR
TI  - Deep3DCCS: Geometry-Driven Collision Cross Section Prediction from Multi-View Molecular Projection Tensors.
AU  - Pattipong Wisanpitayakorn
AU  - Vivek Bhakta Mathema
AU  - Kwanjeera Wanichthanarak
AU  - Thanyanatthamon Keawhawong
AU  - Narumol Jariyasopit
AU  - Pairash Saiviroonporn
AU  - Trongtum Tongdee
AU  - Yongyut Sirivatanauksorn
AU  - Sakda Khoomrung
PY  - 2026
DA  - 2026/09/01
JO  - Journal of chemical information and modeling
DO  - 10.1021/acs.jcim.6c00371
UR  - https://europepmc.org/article/MED/42906041
AB  - Collision cross section (CCS) measurements from ion mobility-mass spectrometry improve confidence in molecular identification in metabolomics and lipidomics and reduce false-positive annotations in untargeted analyses. However, experimental CCS coverage remains incomplete, and accurate computational prediction remains needed. Here, we present Deep3DCCS, a geometry-driven deep learning framework that predicts CCS from stacked multi-angle molecular projection tensors using a three-dimensional convolutional neural network operating across the projection-view and two spatial image dimensions. Three-dimensional molecular structures generated from SMILES strings were rotated around the x, y, and z-axes, converted into binary two-dimensional projections, and assembled into standardized multi-view tensors for CCS regression. Deep3DCCS was developed and evaluated using curated AllCCS and METLIN lipid CCS datasets containing 3948 molecule-adduct entries across [M - H]-, [M + H]+, and [M + Na]+ ions, and input optimization identified 32 × 32 pixel projections with five rotations per axis as a practical balance between prediction accuracy, model stability, and computational feasibility. In independent true-validation experiments, AllCCS-trained models achieved MRPE values of 3.27-3.79% on AllCCS validation data and 2.93-3.72% on METLIN validation data, whereas METLIN-trained models transferred less effectively to the broader AllCCS chemical space. Overall, these results support multi-view molecular projection tensors as a viable geometry-driven representation for CCS prediction and demonstrate that substantial CCS-relevant information can be learned directly from three-dimensional molecular morphology using a 3DCNN architecture, with low single-digit prediction errors across independent and cross-dataset validation experiments.
SN  - 1549-9596
SN  - 1549-960X
ER  -