Pablo Mier

The paper develops a computational framework to predict cooperative and competitive interactions among protein triplets in the human interactome using hyperbolic embeddings and machine learning.

The paper reanalyzes data from Rosenberg et al. to show that codon usage has no significant effect on dihedral angles in protein structures.

The paper uses machine learning to identify functional residues in odorant receptors across insect species, revealing conservation patterns linked to social behavior.

This paper embeds the human protein interaction network in hyperbolic space to reveal latent geometric properties that correlate with biological features and predict protein interactions.