Exploring Evolutionary Fitness in Biological Systems Using Machine Learning Methods

Abstract

Here, we propose a computational approach to explore evolutionary fitness in complex biological systems based on empirical data using artificial neural networks. The essence of our approach is the following. We first introduce a ranking order of inherited elements (behavioral strategies or/and life history traits) in considered self-reproducing systems: we use available empirical information on selective advantages of such elements. Next, we introduce evolutionary fitness, which is formally described as a certain function reflecting the introduced ranking order. Then, we approximate fitness in the space of key parameters using a Taylor expansion. To estimate the coefficients in the Taylor expansion, we utilize artificial neural networks: we construct a surface to separate the domains of superior and interior ranking of pair inherited elements in the space of parameters. Finally, we use the obtained approximation of the fitness surface to find the evolutionarily stable (optimal) strategy which maximizes fitness. As an ecologically important study case, we apply our approach to explore the evolutionarily stable diel vertical migration of zooplankton in marine and freshwater ecosystems. Using machine learning we reconstruct the fitness function of herbivorous zooplankton from empirical data and predict the daily trajectory of a dominant species in the northeastern Black Sea.

Modeling Vertical Migrations of Zooplankton Based on Maximizing Fitness

“…Nevertheless, identification of these coefficients is also possible based on classification methods [33,34]. Let us associate an ordered pair of elements (…”

“…As the experience of using various methods shows [34], the greatest effect can be obtained by using neural networks to solve the set problem. The neural networks technology provides greater flexibility of the algorithm with regard to expanding the training set, adding new experimental results of pair comparison.…”

“…The formulated problem is also a special case of the pattern recognition problem [33,34]. But in contrast to classical problems of this type, here it is not a simple assignment of an element to one of the two classes, but a comparison of elements according to the principle "better or worse".…”

“…For its solution, there is a wide arsenal of computer methods, in particular, machine learning methods (learning-to-rank) [26][27][28][29][30][31][32]. In [33,34], the problem of ranking hereditary elements and identifying the corresponding fitness function was reduced to the problem of classification -dividing ordered pairs of elements into two classes: "the first element is better than the second" and "the second element is better than the first".…”

Towards the next generation of species delimitation methods: an overview of Machine Learning applications

“…Despite some constraints, ML algorithms perform as well as or even outperform model selection methods like ABC and coalescent-based methods (Pei et al, 2018;Smith & Carstens, 2020;Perez et al, 2021;Derkarabetian et al, 2021). Moreover, they are computationally more efficient and generally can be trained on models that are at times too intricate for formal statistical estimators (Pei et al, 2018;Kuzenkov et al, 2020;Smith & Carstens, 2020;Suvorov et al, 2020;Martin et al, 2021;Perez et al, 2021).…”