This repository contains some selected example notebooks and discussions for the analysis of neural data using hidden Markov models on publically accessible data from crcns.org.

All the results shown in these notebooks can be (re-)generated using the accompanying code in this repository, along with publically accessible data from the hc-3 dataset on crcns.org. However, active development has shifted away from this repository, and we are currently actively developing nelpy to make it easier to do common (yet complex) analysis tasks with neuroelectrophysiology data.

Here we demonstrate a few ways in which hidden Markov models (HMMs) can be used to model and analyze sequences of neural activity. This collection is a work in progress, and some of the discussions in these notebooks reflect some of our earlier views and/or approaches. Nevertheless, this collection of notebooks can serve as a good springboard to start thinking about HMMs for neural data analysis.

The easiest way to use this repository is to take a look at the various Jupyter notebooks. These are all the files ending with the .ipynb extensions. These files can be rendered directly in github, so you can look at the code, analysis, and results, all in your browser, without the need to install any additional software, or to execute any code.

The notebooks of particular interest, and in a loosely logical order, are briefly described below.

• StateClustering.ipynb—demonstration that HMMs can (robustly) learn place fields in the absence of positional data; several follow-up analyses showing how behavioral states, different contexts, etc. are captured and clustered in the inferred model states (status: in progress, first half complete)
• StateOrdering.ipynb—post-ordering hidden states for interpretation and evaulation (status: complete)
• ModelSelection.ipynb—sensitivity analysis and discussion on selecting appropriate model parameters (status: complete)
• KLscore.ipynb—preliminary results and proof-of-concept for a Kullback-Leibler based sequence score (status: almost complete)
• SubstateDecoding.ipynb—improving decoding accuracy by considering a substate decoding approach (status: complete)
• EMBC16.ipynb—code needed to generate all the figures from "Scoring sequences of hippocampal activity using hidden Markov models" presented at the 2016 38th Annual International Conference of the Engineering in Medicine and Biology Society (EMBC) (status: complete)

Of course, you can also clone or download this repository, which will allow you to modify the code to see the effect of changing parameters, and so on. For this, you would need Python 3, and several additional packages. Following any god Python tutorial should help you get up and running pretty fast.

The data used here is publically available free-of-charge at http://crcns.org, the data sharing website funded by the Collaborative Research in Computational Neuroscience (CRCNS) program. More specifically we have used dataset hc-3 here, and from that dataset, this repository shows results for animal gor01.

Developed by members of the Kemere Lab at Rice University. This work was supported by the National Science Foundation (CBET-1351692 and IOS-1550994) and the Human Frontiers Science Program (RGY0088). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.