Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 16: Scanning Probe Techniques: Method Development
O 16.11: Vortrag
Montag, 17. März 2025, 17:30–17:45, H25
Molecular Identification via Molecular Fingerprint extraction from Atomic Force Microscopy images — •Manuel González Lastre1, Pablo Pou1, 2, Miguel Wiche3, 4, Daniel Ebeling3, 4, Andre Schirmeisen3, 4, and Rubén Pérez1, 2 — 1Departamento de Física Teórica de la Materia Condensada, Universidad Autónoma de Madrid, E-28049 Spain — 2Condensed Matter Physics Center (IFIMAC), Universidad Autónoma de Madrid, E-28049 Madrid, Spain — 3Institute of Applied Physics, Justus Liebig University Giessen, Giessen, Germany — 4Center for Materials Research, Justus Liebig University Giessen, Giessen, Germany
Previous works have already shown that deep learning (DL) models can retrieve the chemical and structural information encoded in a 3D stack of constant-height HR--AFM images, leading to molecular identification.
In this work, we overcome their limitations by using a well-established description of the molecular structure in terms of topological fingerprints, the Extended Connectivity Fingerprints, which provide local structural information of the molecule. In this work, we train a DL model to extract this optimized structural descriptor from the 3D HR--AFM stacks and use it, through virtual screening, to identify molecules from their predicted ECFP4 with a retrieval accuracy on theoretical images of 95.4%. This approach, unlike previous DL models, assigns a confidence score, the Tanimoto similarity, to each of the candidate molecules, thus providing information on the reliability of the identification.
Keywords: Atomic force microscopy; Molecular identification; Chemical characterization; Neural networks; Molecular fingerprints