Of mice and (wo)men – or: how I learned to read spikes

By Sonja Förster

What are ephys alignments – and why would you do them?

Ephys alignments, or histology – electrophysiology alignments in full, intend to align electrophysiology features of a neural recording to an anatomical histology reference (here, the Allen adult mouse brain atlas) to allow for reasonable alignment of the recording channels along the probe trajectory. This is important since mouse brains might be smaller or larger, or slightly differ in some structural landmarks. And it may be especially relevant for recordings from a Neuropixels probe (or other probes that record across multiple brain regions). Those ephys alignments are therefore a crucial prerequisite to ensure that subsequent analyses compare apples to apples, or, let’s say, spike count variability in a given region of one mouse to the spike count variability in exactly that same region in another mouse.

A discovery in two parts

In spring 2023, I had the opportunity to do my research internship as part of the Resarch Master program Cognitive Neuroscience at Dr Anne Urai’s Cognitive, computational and systems neuroscience – short: CoCoSys lab at Leiden University in the Netherlands. My internship’s core task: do the ephys alignments on Anne’s recordings of aged mice performing a perceptual decision-making task. While during my internship I started my ephys alignment journey on the so-called repeated site recordings (IBL verbiage, referring to the brain regions involved in the IBL reproducability study: Visual Area A (VISa); Hippocampal Field CA1; Dentate Gyrus (DG); Lateral Posterior nucleus of the thalamus (LP); Posterior nucleus of the thalamus (PO)), thereafter, I was able to join Anne’s lab as an RA and to continue the work. This time, aligning what we came to call the non-repeated-site insertions: Most of these trajectories run through more frontal regions as well as several striatal nuclei involved in the integration of visual, motor, and decision-making information.

Ephys alignments – a personal reflection in a nutshell

Doing ephys alignments feels like a treasure hunt to me. Like a puzzle, for which the pieces need to fit and click to eventually comprise this clear full picture. With a slight variation of this metaphor for a beginner in the ephys alignment world: first, identify the pieces which belong to this specific puzzle, to then: click, clear full picture – you know.

Turns out, though, this picture doesn’t always evolve as clearly nor fully as I’d expected when I embarked on the journey to reading spikes. In the beginning, I thought it’s all about learning each region’s ephys features: high or lower firing rates, representing bursts or longer silent phases, rather; high or lower amplitude in spikes; activity in dedicated frequency bands; all of the above in reasonable dependecy to the task performed. And then, when knowing the truth about those patterns, and the related histological landmarks those patterns should fall into, it would only be about the identification of both to be able to align one to the other. In one mouse – and then in all animals: Same features, same landmarks, identify across anmials, align. It didn’t take long for me to realize: This is not how it works. Instead, it is about embracing the uniqueness of each recording while identifying what is shared between them. It is about appreciating that we record from living animals who do many things – task-related and task-unrelated, and at different levels of arousal and engagement, throughout long sessions and many trial repetitions. It is about acknowledging ambiguity, reviewer biases, and the human tendency to take for the truth what we can record and explain and to ignore what we can’t. So, if asked for my biggest learning: It is about the mindset in which to approach this task. It is about doing the puzzle back and forth, aiming to reveal the story in each recording. It is about continuous learning and the willingness to go back and do again as the learning happens. And it is about to not pretend precision and un-ambiguity where it may just not be.

More on the treasure hunt

If you’re interested to read more about my experiences with ephys alignments, my approach and personal logbook of sanity checks, as well as some overall guidelines to get started, you can read the full story starting with the Introduction and project overview or directly jump to any of the subsections

• How I came to terms with the IBL ephys alignment GUI
• A personal account of my biggest learnings in the Overall approach to ephys alignments
• A somewhat more content-focused account of Common ephys features in the mouse brain (IBL repeated site) including my personal logbook of sanity checks and navigating the IBL ephys GUI.

As part of my RA, I also prepared a little presentation as an introduction to ephys alignments for one of our lab meetings. In case you’re interested, here are the slides.