I'm David Silva Sánchez. A PhD student in Applied Mathematics at Yale University. This is a compilation of software development projects related to cryo-EM where I have collaborated or acted as the leading developer.

• Cryo-EM Ensemble Optimization with cryoJAX (Main developer): A physics-guided ensemble optimization method for fitting ensembles of atomic structures to single-particle cryo-EM images. Built with cryoJAX and OpeMM. Details on implementation and experiments available in our preprint.

• Cryo-EM Heterogeneity Challenge (Main developer): Preprocessing and Analysis pipeline for the Cryo-EM Heterogeneity Challenge. Developed the preprocessing pipeline and the PCA metric. A preprint describing the challenge and the results can be in bioarxiv.

• cryoJAX (Collaborator): A flexible and differentiable simulator for cryo-EM data from volumes or atomic structures. Generates data compatible with Relion / Cryosparc. Developed the interface for atom-based volume representations and the interface for reading PDB files and Relion datasets. I am also a main developer for cryoSPAX, a module of cryoJAX that provides tools for developing cryo-EM single-particle analysis pipelines.

• cryoSBI (Collaborator): A simulation-based inference pipeline for estimating the posterior probability of conformational structures given cryo-EM data. Set the basis for the initial design of the API, and collaborated in testing/developing the current API. Results available in our PNAS article.

• AnaVox. Adapted the PCA pipeline from the Cryo-EM Challenge for the analysis of volume series obtained by different heterogeneity methods. See our Nature Communications publication for results.

• Observation on Kmeans: Theoretical study on the behaviour of Lloyd K-Means in the high noise, high dimension regime for a simple k=2 GMM Model. A description of the theory and numerical experiments can be found in our manuscript in ArXiv.