Abstract:The problem of deciphering how low-level patterns (action potentials in the brain, amino acids in a protein, etc.) drive high-level biological features (sensorimotor behavior, enzymatic function) represents the central challenge of quantitative biology. The lack of general methods for doing so from the size of datasets that can be collected experimentally severely limits our understanding of the biological world. For example, in neuroscience, some sensory and motor codes have been shown to consist of precisely timed multi-spike patterns. However, the combinatorial complexity of such pattern codes have precluded development of methods for their comprehensive analysis. Thus, just as it is hard to predict a protein's function based on its sequence, we still do not understand how to accurately predict an organism's behavior based on neural activity. Here we derive a method for solving this class of problems. We demonstrate its utility in an application to neural data, detecting precisely timed spike patterns that code for specific motor behaviors in a songbird vocal system. Our method detects such codewords with an arbitrary number of spikes, does so from small data sets, and accounts for dependencies in occurrences of codewords. Detecting such dictionaries of important spike patterns --- rather than merely identifying the timescale on which such patterns exist, as in some prior approaches --- opens the door for understanding fine motor control and the neural bases of sensorimotor learning in animals. For example, for the first time, we identify differences in encoding motor exploration versus typical behavior. Crucially, our method can be used not only for analysis of neural systems, but also for understanding the structure of correlations in other biological and nonbiological datasets.

Siamese networks for Poincaré embeddings and the reconstruction of evolutionary trees

Assessing similarities between spatial point patterns with a Siamese neural network discriminant model

In the Light of Deep Coalescence: Revisiting Trees Within Networks

Reconstruction of Tree Network Via Evolutionary Game Data Analysis

Identification of quantum entanglement with Siamese convolutional neural networks and semi-supervised learning

Common Variable Learning and Invariant Representation Learning using Siamese Neural Networks

TINNiK: inference of the tree of blobs of a species network under the coalescent model

Siamese Reconstruction Network: Accurate Image Reconstruction from Human Brain Activity by Learning to Compare

Unsupervised Bayesian Ising Approximation for revealing the neural dictionary in songbirds

Ultrafast learning of 4-node hybridization cycles in phylogenetic networks using algebraic invariants

Exploring Simple Siamese Representation Learning

Embedded Spectral Descriptors: Learning the point-wise correspondence metric via Siamese neural networks

Hierarchical-taxonomy-aware and attentional convolutional neural networks for acoustic identification of bird species: A phylogenetic perspective

A Conceptual Network-Based Approach to Inferring the Cultural Evolutionary History of the Baltic Psaltery

Identifiability of Level-1 Species Networks from Gene Tree Quartets

Novel Symmetry-preserving Neural Network Model for Phylogenetic Inference

Learning Evolved Combinatorial Symbols with a Neuro-symbolic Generative Model

Siamese network for classification of Raman spectroscopy with inter-instrument variation for biological applications

Unsupervised Spoken Term Discovery Based on Re-clustering of Hypothesized Speech Segments with Siamese and Triplet Networks

Semi-Supervised Learning using Siamese Networks

Inferring phylogenetic networks from multifurcating trees via cherry picking and machine learning