Our vision

Neuron membrane potentials, being it spikes or graded potentials recorded in dendritic and soma structures or action potentials in axonal segments, are basic biophysical observables at single cell level which determine the physiological functions and behaviors of any multicellular nervous systems.

At Abinitio Laboratories we attempt to understand the critical biophysical and computational mechanisms, in terms of activities of the carrier and ionic channels in cytomembrane, for the creation (how a local cross-membrane electric potential is generated in response to extracellular or intracellular events), propagation (how an electric event propagates from one place to other places on the cytomembrane), integration (how multiple electric signals interact with each other at any loci of cytomembrane), and transformation (how electric and chemical events interrelate with each other) of membrane potentials in dendrites, somas, axons, and presynaptic terminals of a single neuron. 

We take a Computational Neuroelectrophysiology (CNEP) approach to tackling the challenge. In the core of CNEP, we are developing a biophysical model for synergetic interactions amongst LGIC (ligand-gated ion channels), VGIC (voltage-gated ion channels), and DGIC (dual-gated ion channels) in cytomembrane.

For the first time in history of neuroelectrophysiology, CNEP makes it possible to establish a whole-cell, in vivo, end-to-end (from postsynaptic densities, dendritic structures, soma and hillock, axon, through presynaptic terminals), quantitative reconstruction of membrane potential of a single neuron. This represents an ab initio reconstruction of membrane potentials from activities of all participating ion channels, a significant advance in our understanding about how single neurons represent and process information. 

Upon this foundation at single cell level, CNEP further allows an ab initio reconstruction of the signaling process at level of local cortical circuits consisting of tens of thousands of different types of neurons with known afferents and interconnectivity. While it is well established that local neuronal circuits conduct highly specialized physiological functions in sensory perception, memory, and advanced cognitive functions, knowledge remains rare about the underlying cellular and molecular mechanisms due to the difficulty of in vivo simultaneous recording of activities of a large number of neurons. The unique power of CNEP in whole-cell, end-to-end quantitative reconstruction of single neuron’s signal processing offers a computational alternative. This capability would greatly help to decipher the causality of local neuronal circuits with unprecedented details. 

We believe that our work represents a breakthrough in advancing our knowledge about the principles of neuronal computing. It also serves as a disruptive reconstruction of today's artificial intelligence for its building blocks, i.e. the 'neurons' that compute. According to this vision, we are planning to turn CNEP into industry products or solutions that could help neuroscientists, clinic professionals, and AI developers in tackling their challenges.