Maximal Mutual Information Component Analysis
The details on the algorithm and application can be found in Detecting Microbial Dysbiosis Associated with Pediatric Crohn Disease Despite the High Variability of the Gut Microbiota published on Cell Reports in 2016.
Users need to write an IO to input the micrbiome data into the code and analysis the output results. The input data should be a numpy
array with size [N_OTUs, N_samples]
, and the first Nc
samples are control ones. The code for MMICA was written as functions which can be easily understood. To do it, one can implement the function readCell()
to convert the microbiome data in biom
-format or other formats to a numpy
array.
misfunc
includes a function showarray()
, which prints array in a better look. It is not a part of MMICA, and one can implement showarray()
to be simply as
def showarray(x):
print x
- python >= 2.7
- scipy
Please consider to cite
@article{wang_detecting_2016,
title = {Detecting {Microbial} {Dysbiosis} {Associated} with {Pediatric} {Crohn} {Disease} {Despite} the {High} {Variability} of the {Gut} {Microbiota}},
volume = {14},
url = {http://www.sciencedirect.com/science/article/pii/S2211124715015442},
doi = {10.1016/j.celrep.2015.12.088},
number = {4},
journal = {Cell Reports},
author = {Wang, Feng and Kaplan, Jess L. and Gold, Benjamin D. and Bhasin, Manoj K. and Ward, Naomi L. and Kellermayer, Richard and Kirschner, Barbara S. and Heyman, Melvin B. and Dowd, Scot E. and Cox, Stephen B. and Dogan, Haluk and Steven, Blaire and Ferry, George D. and Cohen, Stanley A. and Baldassano, Robert N. and Moran, Christopher J. and Garnett, Elizabeth A. and Drake, Lauren and Otu, Hasan H. and Mirny, Leonid A. and Libermann, Towia A. and Winter, Harland S. and Korolev, Kirill S.},
month = feb,
year = {2016},
pages = {945--955}
}