Model a given k-mer profile as a mixture of multiple evenly spaced negative binomial distributions, and fit parameters by minimizing the residual sum of squares errors.
The model used here is a generalized version of the GenomeScope model, extended to allow explicit modelling of errors and fitting an arbitrary number of peaks.
pip install git+https://github.com/jzhang-dev/kmer-profile-decomposer
See demo.py.ipynb.
Vurture, G. W., Sedlazeck, F. J., Nattestad, M., Underwood, C. J., Fang, H., Gurtowski, J., & Schatz, M. C. (2017). GenomeScope: fast reference-free genome profiling from short reads. Bioinformatics, 33(14), 2202-2204.