Machine learning has greatly expanded the ability to classify images. However, many machine learning classifiers require thousands of images for training and lack quantitative descriptors of how images were grouped. We overcome these limitations with a machine learning approach based on topological data analysis, where a data set of 20-30 images is sufficient to accurately train the classifier. Our method quantifies differences between groups and identifies subcellular regions with the largest dissimilarities.
29 June 2026 in Journal of Applied and Computational Topology
We construct weak homotopy equivalences between the geometric realizations of directed Vietoris-Rips complexes and their underlying directed graphs, seen as pseudotopological spaces. Pseudotopological spaces are a generalization of (Čech) closure spaces which in turn generalize topological spaces, but they also include graphs and directed graphs as full subcategories, making them a natural bridge that connects classical algebraic topology with the more applied side of topology. This weak homotopy equivalence implies that singular homology groups of finite directed graphs can be efficiently calculated from finite combinatorial structures, despite their associated chain groups being infinite dimensional. Along the way, we establish analogues of classical results such as the existence of a long exact sequence for homotopy groups of pairs of pseudotopological spaces and that a weak homotopy equivalence induces isomorphisms for homology groups. This work is similar to the work of McCord for finite topological spaces but in the context of pseudotopological spaces. Our results also give a novel approach for studying (higher) homotopy groups of discrete mathematical structures such as (directed) graphs or digital images.
Representations of chemical features in the neural activity of the olfactory bulb (OB) are not well-understood, unlike the neural code for stimuli of the other sensory modalities. This is because the space of olfactory stimuli lacks a natural coordinate system, and this significantly complicates characterizing neural receptive fields (tuning curves), analogous to those in the other sensory modalities. The degree to which olfactory tuning is spatially organized across the OB, often referred to as chemotopy, is also not well-understood. To advance our understanding of these aspects of olfactory coding, we introduce an interpretable method of Shapley fields, as an olfactory analog of retinotopic receptive fields. Shapley fields are spatial distributions of chemical feature importance for the tuning of OB glomeruli. We used this tool to investigate chemotopy in the OB with diverse sets of chemical features using widefield epifluorescence recordings of the mouse dorsal OB in response to stimuli across a wide range of the chemical space. We found that Shapley fields reveal a weak chemotopic organization of the chemical feature sensitivity of dorsal OB glomeruli. This organization was consistent across animals and mostly agreed across very different chemical feature sets: (i) the expert-curated PubChem database features and (ii) features derived from a Graph Neural Network trained on human olfactory perceptual tasks. Moreover, we found that the principal components of the Shapley fields often corresponded to single commonly accepted chemical classification groups, that therefore could be “recovered” from the neural activity in the mouse OB. Our findings suggest that Shapley fields may serve as a chemical feature-agnostic method for investigating olfactory perception.
11 December 2024 in SIAM Journal on Applied Mathematics
We describe the combinatorics of equilibria and steady states of neurons in threshold-linear networks that satisfy Dale’s law. The combinatorial code of a Dale network is characterized in terms of two conditions: (i) a condition on the network connectivity graph, and (ii) a spectral condition on the synaptic matrix. We find that in the weak coupling regime the combinatorial code depends only on the connectivity graph, and not on the particulars of the synaptic strengths. Moreover, we prove that the combinatorial code of a weakly coupled network is a sublattice, and we provide a learning rule for encoding a sublattice in a weakly coupled excitatory network. In the strong coupling regime we prove that the combinatorial code of a generic Dale network is intersection-complete and is therefore a convex code, as is common in some sensory systems in the brain.