Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 32 for “"Molecular Graphs"”.
-
Independence Polynomials of Molecular Graphs
In the 1980's, it was noticed by molecular chemists that the stability and boiling point of certain molecules were related to the number of independent vertex sets in the molecular graphs of those chemicals. This led to the definition of the Merrifield-Simmons index of a graph G as the number of …
-
Deep Learning for Drug Discovery, Property Prediction with Neural Networks on Raw Molecular Graphs
… network architectures learning directly on raw molecular graphs, thereby eliminating the feature engineering step. The starting point of experimentation is a reimplementation of the previously proposed message passing neural networks framework for learning on graphs, analogous to convolutional …
-
Machine Learning-Based Prediction of Sites of Metabolism in Drugs: Exploring Feature Extraction Methods on Molecular Graphs
… tasks. Seeing molecules as attributed graphs, where atoms correspond to nodes and bonds correspond to edges, the SOM prediction problem can be formulated as a node classification task. We compare two methods of extracting node features from molecular graphs: a standard fingerprint …
-
Graph representation learning: algorithm development and its application to early drug discovery
… powerful capability to flexibly handle various molecular and biological interaction information at different data scales, enabling effective reduction of the search space of following wet experiment validations based on efficient candidate screenings. Focusing on one of the most popular GRL …
-
Transferable User-Friendly Machine Learning for Normal Boiling Point Prediction
… relationships (QSPR), which aim to correlate molecular descriptors with various features of the substance. Although in recent years, Machine Learning (ML) and Deep Learning (DL) techniques have gained traction for predicting molecular properties, determining the optimal strategies for feature …
-
Deep Learning for Molecular Geometry and Property Analysis
… multiple layers to model complex patterns in molecular data. By leveraging large datasets and powerful computational capabilities, deep learning is revolutionizing molecular sci-ence, enabling faster, more accurate predictions and discoveries that were previously challenging or impossible. …
-
Structure-guided Genetics-informed Embeddings Enable Zero-shot Drug Sensitivity Prediction
… that learns drug representations directly from molecular structure, supervised by pharmacogenomic similarity derived from response profiles. By aligning structural embeddings with biological response similarity, the model captures meaningful relationships between drugs while remaining …
-
Machine Learning for Predicting Targeted Protein Degradation
… to leverage different PROTAC representations: molecular fingerprints, molecular graphs and tokenized SMILES. The proposed models are evaluated against an XGBoost model baseline and the State-of-The-Art (SOTA) model for predicting PROTACs degradation activity. Overall, our best DL models …
-
Motif-based Graph Neural Networks: Representation, interpretation, and extraction
… (GNNs) have emerged as powerful tools for molecular representation learning, but they often struggle to capture higher-order structural properties and suffer from a lack of transparency. This thesis addresses these limitations by leveraging motifs—recurring and statistically significant …
-
Graph Representation Learning of Molecules for de novo Drug Development
… have provided valuable tools for predicting molecular properties, but these methods can fall short in terms of efficiency and predictive accuracy. When used in molecular generation, these limitations become more pronounced, as inaccuracies in property prediction can lead to the design of …
-
The Mechanisms of Amide Hydrolysis
… state to the reactants and products. All of the molecular geometries were obtained using the 4-31G basis set and all wavefunctions and energies were calculated at the 6-31G** level of theory. The theory of atoms in molecules was applied to each reaction to study the mechanisms of structural …
-
Investigating graph neural networks and classical feature-extraction techniques in activity-cliff and molecular property prediction
<p>Molecular featurisation refers to the process of transforming molecular data into numerical feature vectors. It is one of the key research areas in molecular machine learning and computational drug discovery. Recently, message-passing graph neural networks (GNNs) have emerged as a novel method …
-
Hierarchical Language Models for Graphs: Towards Better Performance and Interpretability
… fail to grasp the complex structural details of graphs and struggle with basic tasks such as identifying node degrees or determining the shortest paths between nodes. Moreover, even small graphs can have lengthy natural language descriptions, resulting in scalability issues and making it …
-
Gaussianization of the spectra of graphs and networks : theory and applications
… understanding important structural properties of graphs and networks, such as communicability, node centrality, bipartivity and many more. Here we propose a new matrix function based on the Gaussianization of the adjacency matrix of a graph. This function gives more weight to a selected reference …
-
Application of Explainable Graph Neural Networks in Drug and Biomarker Discovery for Human Diseases
… learning capabilities across both structural graphs (e.g., molecular graphs) and non-structural graphs (e.g., patient similarity networks). This dissertation presents cross-disciplinary methodological and translational research that develops structure-aware and knowledgeaware graph AI …
-
Unsupervised Learning of Molecular Representations for Drug Development
… of interest in using machine learning to predict molecular properties such as the binding affinity to protein targets of interest or to generate molecular structures with desirable properties. However, as chemical entities are challenging to represent in an expressive and computer-interpretable …
-
Geometric Deep Learning for Science: Prediction, Generation, and Symmetries
… need to be dramatically modified to suit for 2D molecular graphs, 3D molecular geometries, and periodic materials. In this dissertation, we study these challenges and propose several novel methods to tackle them. We first propose GraphAug, a novel automated data augmentation method aiming at …
-
Graph Neural Networks: From Expressivity to Generalisation
Graphs are mathematical representations widely used to describe complex relationships in various fields. Graph machine learning techniques, such as Graph Neural Networks (GNNs), have recently demonstrated remarkable success in extracting meaningful patterns from intricate graph data, enabling …
-
Learning from Structures: Graphs, Knowledge, and Judgement
… can be effectively modeled using formats such as graphs, tables, and hierarchical schemes. For example, molecular graphs are foundational in drug discovery, protein contact maps encode 3D biological interactions, social networks reflect relational patterns in communication, and logic chains …
-
Learning from Structures: Graphs, Knowledge, and Judgement
… can be effectively modeled using formats such as graphs, tables, and hierarchical schemes. For example, molecular graphs are foundational in drug discovery, protein contact maps encode 3D biological interactions, social networks reflect relational patterns in communication, and logic chains …
Page 1 of 2