Robust and Scalable Models of Microbiome Dynamics

Authors: Travis Gibson, Georg Gerber

ICML 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We apply our method to simulated and real data, and demonstrate the utility of our technique for system identification from limited data, and for gaining new biological insights into bacteriotherapy design.
Researcher Affiliation Academia Travis E. Gibson 1 Georg K. Gerber 1 1Massachusetts Host-Microbiome Center, Brigham and Women s Hospital, Harvard Medical School, Boston, MA, USA. Correspondence to: TE Gibson <tgibson@mit.edu>, GK Gerber <ggerber@bwh.harvard.edu>.
Pseudocode No The paper describes the inference algorithm and model components in text and with a graphical model (Figure 2), but does not include structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide an explicit statement about releasing its source code or a direct link to a code repository for the described methodology.
Open Datasets Yes We next applied our algorithm to real data from (Bucci et al., 2016), which investigated developing a bacteriotherapy for Clostridium difficile...
Dataset Splits No While the paper describes using 'simulated data' and 'real data from (Bucci et al., 2016)' with some details on replicates and time points, it does not provide explicit training/test/validation split percentages, sample counts, or specific predefined split citations necessary for reproduction.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory specifications, or detailed computer configurations) used for running its experiments.
Software Dependencies No The paper mentions statistical models and methods used (e.g., Negative Binomial distribution, Bayesian adaptive lasso, references to DESeq2), but does not provide specific software dependencies with version numbers (e.g., 'Python 3.8, PyTorch 1.9').
Experiment Setup Yes To simulate these experiments, we generated data with 5 biological replicates (5 different time series simulated from the same dynamics, but with different initial conditions), 11 time-points per replicate... Our model found a median of 4 interaction modules (5,000 MCMC samples with 1,000 burnin).