Software engineer and former neuroscientist working on new tools for research

Experienced programmer with advanced ability in several languages (MATLAB, Python, C#, JavaScript/HTML/CSS)

Hardware design and prototyping with focus on custom equipment for behavioural neurophysiology. Proficient in 3D modeling (AutoDesk Inventor)

Extensive experience with data analysis approaches for neural and behavioural data

WORK

2022 - Present

Software Engineer NeuroGEARS Ltd.

Developing and extending the Bonsai-Rx ecosystem.

2018 - 2022

Postdoctoral Scholar - Hires lab, University of Southern California

Utilising all-optical techniques to investigate neuronal ensemble recruitment in somatosensory cortex. Using 2-photon imaging with spatial-light modulation optogenetics to investigate information propagation across cortical layers, and the effects of ensemble recruitment on tactile perception.

Modified DeepLabCut for Google Cloud - allowing for fast, parallel usage on large whisker tracking datasets.

2013 - 2018

PhD Student - Schaefer lab, The Francis Crick Institute / UCL

Developed an automated, high-throughput mouse behavior system (AutonoMouse) capable of training 20+ mice simultaneously in GNG/2AFC paradigms; as well as the associated control and analysis software.

Utilised AutonoMouse in a large-scale study investigating the effect of targeted olfactory bulb (OB) lesions on olfactory discrimination behaviors (Erskine 2019, PLOS ONE)

Using a custom-built odor delivery device, in vivo electrophysiology and automated behavior: investigated the role of temporal patterns in odor concentration on detection and coding in the mouse OB (Ackels, T., Erskine, A. Nature, accepted).

Produced several software solutions for use in the lab (AutonoMouse, PulseBoy).

2012 - 2013

Master's Project Student - Petersen lab, The University of Manchester

Developed protocol and experimental rig for head-fixed GNG behavioral training in mice with simultaneous awake in vivo electrophysiology and high-speed recording / tracking of whisker movement.

Using this approach, investigated the coding of whisker kinematics in awake mice at the primary afferent level (Campagner et al., 2016; Bale et al., 2015)

Summer 2012

Visiting Scholar - Paninski Lab, Columbia University

Project investigating approaches for decoding limb position from macaque PMd multi-site recordings for neural prosthetic applications.

Summer 2011

Summer Placement Student - Petersen Lab, The University of Manchester

Project investigating encoding of whisker stimulation in rodent somatosensation. Built and tested a number of encoding models accounting for whisker afferent firing under white-noise stimulation.

PUBLICATIONS

Fast odour dynamics are encoded in the olfactory system and guide behaviour

Ackels, T.*, Erskine, A.*, Dasgupta, D.*, Cristina, M., Warner, T., Tootoonian, S., Fukunaga, I., Harris, J., Schaefer, A. T. (Nature, 2021)

In this project we investigated whether the turbulent dynamics of naturally occurring odours were a useful sensory component of animal behavior. Ultimately we found that mice were able to discriminate correlation structure in odour stimuli at high frequency, suggesting they are capable of using turbulent odour signals for navigation and localisation. We overcame several large technical hurdles to achieve this finding, including development of a new method of measuring different odours simultaneously in the air (defPID); a new olfactometer design capable of extremely precise and fast odour delivery; and integration of these tools into our previously developed automated training platform: AutonoMouse.

Behavioral and Neural Bases of Tactile Shape Discrimination Learning in Head-Fixed Mice

Kim, J., Erskine, A., Cheung, J. D., Hires, S. A. (Neuron 2020)

Using 2-photon imaging and head-fixed behavior, we uncovered how S1 neurons encode tactile shape, and how this encoding shifts over the course of learning.

AutonoMouse: High throughput operant conditioning reveals progressive impairment with graded olfactory bulb lesions

Erskine, A., Bus, T., Herb, J. T., Schaefer, A. T. (PLOS ONE 2019)

We developed AutonoMouse to reduce the bottleneck in neuroscience research from manual or semi-manual behavioral training and testing techniques. In AutonoMouse, large cohorts of mice can be trained simultaneously over long periods with minimal experimenter intervention. Experiments in AutonoMouse can be designed, implemented and monitored remotely with dedicated software. Using this system, we performed a systematic study of the effect of olfactory bulb disruption on olfactory discrimination, finding that olfactory memory was the most prominently modified behaviour.

Prediction of primary somatosensory neuron activity during active tactile exploration

Campagner, D., Evans, H. M., Bale, M. R., Erskine, A., Petersen, R. S. (eLife 2016)

This study used high-speed videography with awake electrophysiology to uncover the kinematic variables encoded by trigeminal afferents in the whisker system.

Microsecond-scale timing precision in rodent trigeminal primary afferents

Bale, M. R., Campagner, D. Erskine, A., Petersen, R. S. (JNeurosci 2015)

Here we uncovered the extraordinary precision in spike-timing and information capacity of rodent whisker trigeminal afferents.