1st International and 10th National Iranian Conference on Bioinformatics

Home Accepted Papers

Inferring microbial communities using constrained damped lasso regression based on the generalized Lotka-Volterra model

Paper ID : 1236-ICB10

Authors:

Naser Elmi, Kaveh Kavousi *, Ahmad Kalhor

University of Tehran

Abstract:

In this study, we developed a new method to infer microbial communities. Synthetic and natural microbial communities play essential roles in the industry and our health [1], [2]. We used the generalized Lotka-Volterra (GLV) model to model the interaction between microbes. Due to the meaningfulness of the parameters in this model, it has been widely used for modeling microbial communities [3]. To the best of our knowledge, most of the regression-based methods on the GLV equation did not consider the sparsity and constraints of the real problem. Hence, to solve the limitations of the available methods, we developed damped lasso regularization to solve this constrained-based convex optimization problem. We used CVX solver for this problem [4]. We trained and tested our method on various simulated microbial communities with 3 to 5 interacting microbes with different dynamics, including stable fixed point, limit cycle, and chaotic dynamics. We used the cross-validation method to test our method's performance in inferring the magnitude and sign of the interactions. We calculated the correlation of the estimated abundance of interacting microbes based on the inferred model with actual data. Our results demonstrated that the developed method could accurately predict the parameters sign and magnitude. Furthermore, the correlation between the estimated abundance and real abundance was more than 0.9. We also evaluated the model performance in presence of noise.

Keywords:

microbial community, inference, lasso regression, GLV, Optimization

Status : Paper Accepted (Oral Presentation)