This is the code release for paper: Continuous State HMMs for Modeling Time Series Single Cell RNA-Seq Data

See supporting website:
http://www.andrew.cmu.edu/user/chiehl1/CSHMM/

This project has beed test on Ubuntu 14.04 and Windows 10. This project might or might not work on other platforms.

Please install scdiff (version 1.1.4) for model initialization:
http://www.cs.cmu.edu/~jund/scdiff/index.html

Please install gprofiler 0.2.3 for GO analysis
https://pypi.python.org/pypi/gprofiler-official/0.2.3
Remember to change the path for gprofiler.py in CSHMM_analysis.py
line 360:     
cmd = 'python ~/repos/bio_packages/gprofiler-official-0.2.3/gprofiler.py -o mmusculus "'+ ' '.join(gene_list) + '" >> '+out_file


See the ipython notebook file tutorial.ipynb for an example of using our CSHMM model to analyze time series single cell RNA-Seq data
The lung_results folder contains .txt files (GO analysis) and png files (gene expression figures) for example data treutlein2014

Note that if windows is used, some unix shell script commands are not executable. The GO analysis tables (.txt files) will be lack of headers. The header is as following:
"query	significant	p_value	T	Q	Q&T	precision	recall	term_id	domain	group	description	depth	intersection	evcodes"


This project needs the following:
 python 2.7
 graphviz (Windows: remember to add path variable after installation: C:\Program Files (x86)\Graphviz2.38\bin. Note that this path depends on where you install graphviz)
 
 python packages:
 cvxpy (Windows: Microsoft Visual C++ 9.0 is required. Get it from http://aka.ms/vcpython27 )
 progressbar
 pygraphviz
 matplotlib
 networkx
 numpy
 pandas
 argparse
 sklearn