Ramón Alain Miranda‐Quintana

This study evaluates contrastive learning-based approaches for activity cliff prediction in drug discovery, finding they improve embedding separation but have dataset-dependent accuracy impacts.

This study examines how undersampling techniques affect hyperparameter selection for non-linear dimensionality reduction methods in chemical space visualization, showing reduced computational costs and improved local structure preservation.

The paper improves k-means clustering for MD simulations with new seeding strategies, reducing runtime without sacrificing clustering quality.

The paper introduces an active learning strategy using Thompson sampling to efficiently screen peptide libraries with AlphaFold-based binding predictions, achieving significant improvements over random sampling.

mdBIRCH clusters molecular dynamics trajectories efficiently using an online method with a RMSD-calibrated merge test, enabling scalable and interpretable analysis.

DIVINE provides a deterministic clustering framework for molecular dynamics trajectories, achieving better clustering quality and efficiency than bisecting k-means.

Selector is a Python library for subset selection that provides diverse sampling algorithms and integrates with data analysis tools, with applications in computational chemistry and drug discovery.

PRISM clusters molecular pathways using structural medoids and geometric dissimilarities, showing robust clustering across complex biomolecular transitions.

This paper introduces strategies to optimize BitBIRCH clustering for large molecular libraries by adjusting similarity thresholds and branching factors.

BitBIRCH-Lean efficiently clusters large molecular libraries on workstations, achieving 2X speedup over the original BitBIRCH with no loss in cluster quality.

The paper introduces a method to visualize chemical space by projecting molecular data onto a lower-dimensional space using linear methods, outperforming non-linear techniques in efficiency and neighborhood preservation.

The paper benchmarks undersampling techniques for imbalanced DNA-encoded library datasets using three machine learning models, finding that max_sim performs best and implementing the pipeline in DELight.

The paper introduces methods to identify and avoid activity cliffs in drug design using the BitBIRCH clustering algorithm.

The paper introduces a new density-based clustering algorithm for molecular dynamics data to improve the identification of metastable states and rare events.

The paper analyzes the time evolution of chemical library diversity using molecular fingerprints and finds that increased molecule count does not necessarily equate to increased diversity.

This paper introduces HELM, a hybrid clustering method combining k-means efficiency with hierarchical flexibility, and demonstrates its application to protein-DNA and protein folding simulations.

The paper introduces BitBIRCH clustering refinement strategies to improve clustering quality without sacrificing efficiency, with new post-processing tools and examples.

The paper introduces iCliff, a linear complexity method to quantify activity cliffs, showing strong correlation with SALI-based metrics and identifying key molecules contributing to activity cliffs.

SHINE improves pathway ensemble analysis by integrating n-ary similarity, cheminformatics tools, and hierarchical clustering for MD simulations.

This paper improves Radial Threshold Clustering by reducing its time complexity to O(N) and making clustering results order-invariant.

The paper demonstrates how to generate a metric consistent with Tanimoto similarity and explores its properties and relationships with other similarity measures.

The paper introduces efficient methods to calculate standard deviation of molecular similarities with reduced computational complexity.

The paper introduces BitBIRCH, a clustering algorithm for large molecular libraries that is faster and more memory-efficient than existing methods.

PRIME improves identification of native-like protein structures from MD simulations by using extended similarity indices, achieving lower RMSD than traditional centroid-based methods.

The paper introduces NANI, a deterministic k-means clustering algorithm for molecular dynamics simulations that improves cluster accuracy and reproducibility compared to k-means++.

This paper introduces extended similarity indices to improve the interpretation of imaging mass spectrometry data by complementing PCA and enabling efficient analysis of large datasets.