A mechanism-aware and multiomic machine-learning pipeline characterizes yeast cell growth

Abstract

Metabolic modeling and machine learning are key components in the emerging next generation of systems and synthetic biology tools, targeting the genotype–phenotype–environment relationship. Rather than being used in isolation, it is becoming clear that their value is maximized when they are combined. However, the potential of integrating these two frameworks for omic data augmentation and integration is largely unexplored. We propose, rigorously assess, and compare machine-learning–based data integration techniques, combining gene expression profiles with computationally generated metabolic flux data to predict yeast cell growth. To this end, we create strain-specific metabolic models for 1,143 Saccharomyces cerevisiae mutants and we test 27 machine-learning methods, incorporating state-of-the-art feature selection and multiview learning approaches. We propose a multiview neural network using fluxomic and transcriptomic data, showing that the former increases the predictive accuracy of the latter and reveals functional patterns that are not directly deducible from gene expression alone. We test the proposed neural network on a further 86 strains generated in a different experiment, therefore verifying its robustness to an additional independent dataset. Finally, we show that introducing mechanistic flux features improves the predictions also for knockout strains whose genes were not modeled in the metabolic reconstruction. Our results thus demonstrate that fusing experimental cues with in silico models, based on known biochemistry, can contribute with disjoint information toward biologically informed and interpretable machine learning. Overall, this study provides tools for understanding and manipulating complex phenotypes, increasing both the prediction accuracy and the extent of discernible mechanistic biological insights.

A Hybrid Flux Balance Analysis and Machine Learning Pipeline Elucidates Metabolic Adaptation in Cyanobacteria

“…It can be argued that analyzing single-omic data alone has limited relevance in the context of metabolic processes, as it does not capture the full complexity of the phenotype in relation to environmental variability. The hybrid approach proposed in this work connects transcriptomic and fluxomic data using a data-driven multi-view approach that supports machine learning algorithms to yield more accurate predictions ( Culley et al., 2020 ). Considering the vast dimensionality of multi-omic models, the identification of biologically meaningful information can prove to be challenging.…”

Multimodal regularized linear models with flux balance analysis for mechanistic integration of omics data

“…Multi-Modal artificial Neural Networks (MMNN) are a particular type of ANNs devised for learning from different sources of information, in general involving the use of an independent network for processing each modality and then a further network for integrating the gathered information and producing an output. In order to make a fairer comparison between the RLMs and the neural networks, in this study we trained the architecture devised in (Culley et al, 2020), which inherently works in our scenario. This network is composed of two individual networks (one for the fluxomics data and one for the transcriptomics one) whose outputs are then concatenated and further processed by another separate network.…”

“…The goal is to reveal what characteristics a model should have to take advantage of this heterogeneous information. Despite superior prediction accuracy recently observed for multimodal neural networks (Culley et al, 2020), as noted above there may be several factors hindering their utilisation in some case studies. Moreover, the interest in combining regularised linear methods with fluxomics data could also be motivated by the enhancement of model interpretation in biological terms.…”

“…Previous research focused on building linear predictive models for yeast growth (Airoldi et al, 2009), and more recently machine learning both for E. coli and S. cerevisiae (Wytock and Motter, 2019). Metabolic activity in combination with machine learning techniques was taken into consideration and evaluated only lately (Culley et al, 2020).…”

Constraint-based modeling and machine learning applications for analysis and optimization of fermentation parameters

“…This strategy also has been adapted to predict yeast S. cerevisiae growth rate. In this study, fluxomics, generated from parsimonious flux balance analysis (pFBA), were coupled with transcriptomics to train neural networks [122].…”