Development of an integrated pipeline for correlating and predicting orofacial features and 2-photon neural data in mice

We are interested in studying the cortical circuits involved in sensory processing with a focus on the role of inhibitory interneurons in these circuits.

Sensory perception is shaped by an animal’s internal state. A large portion of neural activity in the cortex is spontaneously active even in the absence of external stimuli. This spontaneous activity is modulated by the animal’s behavior and internal state.

Many behavior patterns such as locomotion, whisking and blinking correlate with spontaneous activity across the sensory areas. We are interested in developing a processing pipeline that will integrate these fine orofacial movements acquired via an infrared camera with the neural activity recording in sensory cortical areas via 2-photon microscopy.

This project will involve using the facemap module (Pachitariu, M. & Stringer, C. (2024)) to build a pipeline that will preprocess, annotate, extract keypoints from facial camera recording of mice undergoing in-vivo calcium recording. Compute and derived metrics such as principal components of the recorded orofacial movements and, using facemap, apply Hidden Markov models to identify the activity sequences of the spontaneous behaviors and also output them into time series data and then correlate and predict neural activity.

Additionally, the student will learn how to perform patch clamp recordings in acute brain slices to assess the strength of specific inputs to neuronal populations in the visual cortex. 

Expected project outcomes:

• Configure Facemap/Suite2p integration pipeline.
• Keypoint labelling and tracking of orofacial features via facemap.
• Optimizing the preprocessing, ROI detection and dF/F extraction via suite2p.
• Synchronization, aligning the facemap, suite2p and locomotion outputs.
• Extract behavior time series.
• Apply HMM for behavioral state classification.
• Prediction of neural activity from facemap features.
• Analyzing, plotting and presenting the combined outputs.
• Learn to patch successfully