BACKGROUND: Microorganisms typically exist in communities, where interactions among them define the complexity of these ecosystems. Developing in silico frameworks to investigate the behavior and functionality of these communities is therefore essential for advancing our understanding of microbial ecology. In recent years, several computational modeling frameworks based on genome-scale models have been developed for the community-level analysis of microbial systems. RESULTS: Here, we introduce microbial optimization without forced altruism (MOFA), a bilevel optimization framework that considers both species-level and community-level fitness criteria. By imposing constraints on species biomass in the outer problem, it prevents the forced altruism observed in previous algorithms. We applied MOFA to a toy model and to community models of Desulfovibrio vulgaris and Methanococcus maripaludis, which exhibit a cross-feeding relationship that causes the community objective to override individual fitness goals by prioritizing the export of metabolites for other community members. For this microbial community, a comparison with the results of NECom, OptCom, and Joint-FBA shows that MOFA yields predictions that better match the experimental results. Additionally, for pairs with a cross-feeding relationship in which exported metabolite production associated with this mutual interaction competes with species biomass, such as D. vulgaris and M. maripaludis, NECom fails to predict the community growth rate, whereas our method succeeds. CONCLUSIONS: MOFA effectively analyzes community growth rates without relying on forced altruism. In cases where NECom fails to predict community growth, MOFA successfully predicts these growth rates. Furthermore, MOFA enhances computational efficiency by eliminating the need for the binary variables required in the NECom algorithm.

MOTIVATION: Metabolic modeling has emerged as a powerful tool for predicting community functions. However, current modeling approaches face significant challenges in balancing the metabolic trade-offs between individual and community-level growth. In this study, we investigated the effect of metabolic relatedness among taxa on growth rate calculations by merging related taxa based on their metabolic similarity, introducing this approach as PhyloCOBRA. RESULTS: This approach enhanced the accuracy and efficiency of microbial community simulations by combining genome-scale metabolic models (GEMs) of closely related organisms, aligning with the concepts of niche differentiation and nestedness theory. To validate our approach, we implemented PhyloCOBRA within the MICOM and OptCom package (creating PhyloMICOM and PhyloOptCom, respectively), and applied it to metagenomic data from 186 individuals and four-species synthetic community (SynCom). Our results demonstrated significant improvement in the accuracy and reliability of growth rate predictions compared to the standard methods. Sensitivity analysis revealed that PhyloMICOM models were more robust to random noise, while Jaccard index calculations showed a reduction in redundancy, highlighting the enhanced specificity of the generated community models. Furthermore, PhyloMICOM reduced the computational complexity, addressing a key concern in microbial community simulations. This approach marks a significant advancement in community-scale metabolic modeling, offering a more stable, efficient, and ecologically relevant tool for simulating and understanding the intricate dynamics of microbial ecosystems. AVAILABILITY AND IMPLEMENTATION: PhyloCOBRA implementations are available as extensions to the MICOM packages and can be accessed at https://github.com/sepideh-mofidifar/PhyloCOBRA.

Radiation therapy, treating over half of all cancer patients, involves using specialized machines to direct high-energy beams at tumors, aiming to damage cancer cells while minimizing harm to nearby healthy tissues. Customizing the shape and intensity of radiation beams for each patient leads to solving large-scale constrained optimization problems that need to be solved within tight clinical time-frame. At the core of these challenges is a large matrix that is commonly sparsified for computational efficiency by neglecting small elements. Such a crude approximation can degrade the quality of treatment, potentially causing unnecessary radiation exposure to healthy tissues—this may lead to significant radiation-induced side effects—or delivering inadequate radiation to the tumor, which is crucial for effective tumor treatment. In this work, we demonstrate, for the first time, that randomized sketch tools can effectively sparsify this matrix without sacrificing treatment quality. We also develop a novel randomized sketch method with desirable theoretical guarantees that outperforms existing techniques in practical application. Beyond developing a novel randomized sketch method, this work emphasizes the potential of harnessing scientific computing tools, crucial in today's big data analysis, to tackle computationally intensive challenges in healthcare. The application of these tools could have a profound impact on the lives of numerous cancer patients. Code and sample data available at https://github.com/PortPy-Project/CompressRTP

Computer-assisted collaborative peer grading is a developing growth area in academic evaluation. However, peer assessment often needs help with problems such as the lack of reliability, transparency, fairness, grading speed, and motivation to participate among students. The literature suggests several principles that each partly address the said issues. We propose a novel decentralized approach to academic peer assessment, using blockchain as an underlying technology, to address the principal problems in traditional peer assessment. We also derive design concepts for a modern courseware (CW) application consisting of our method and apply them to implement our approach in a CW called Blockment. We test the effectiveness of our method and system by running quantitative and qualitative experiments, proving our claims of improving reliability, transparency, fairness, grading speed, and motivation of grades in peer assessment. The results suggest embedding our method and system in academic courses to improve conventional peer grading methods.

Microbial communities play fundamental roles in every complex ecosystem, such as soil, sea and the human body. The stability and diversity of the microbial community depend precisely on the composition of the microbiota. Any change in the composition of these communities affects microbial functions. An important goal of studying the interactions between species is to understand the behavior of microbes and their responses to perturbations. These interactions among species are mediated by the exchange of metabolites within microbial communities. We developed a computational model for the microbial community that has a separate compartment for exchanging metabolites. This model can predict possible metabolites that cause competition, commensalism, and mutual interactions between species within a microbial community. Our constraint-based community metabolic modeling approach provides insights to elucidate the pattern of metabolic interactions for each common metabolite between two microbes. To validate our approach, we used a toy model and a syntrophic co-culture of Desulfovibrio vulgaris and Methanococcus maripaludis, as well as another in co-culture between Geobacter sulfurreducens and Rhodoferax ferrireducens. For a more general evaluation, we applied our algorithm to the honeybee gut microbiome, composed of seven species, and the epiphyte strain Pantoea eucalypti 299R. The epiphyte strain Pe299R has been previously studied and cultured with six different phyllosphere bacteria. Our algorithm successfully predicts metabolites, which imply mutualistic, competitive, or commensal interactions. In contrast to OptCom, MRO, and MICOM algorithms, our COMMA algorithm shows that the potential for competitive interactions between an epiphytic species and Pe299R is not significant. These results are consistent with the experimental measurements of population density and reproductive success of the Pe299R strain.

BACKGROUND: High-throughput omics technologies have enabled the comprehensive reconstructions of genome-scale metabolic networks for many organisms. However, only a subset of reactions is active in each cell which differs from tissue to tissue or from patient to patient. Reconstructing a subnetwork of the generic metabolic network from a provided set of context-specific active reactions is a demanding computational task. RESULTS: We propose SWIFTCC and SWIFTCORE as effective methods for flux consistency checking and the context-specific reconstruction of genome-scale metabolic networks which consistently outperform the previous approaches. CONCLUSIONS: We have derived an approximate greedy algorithm which efficiently scales to increasingly large metabolic networks. SWIFTCORE is freely available for non-commercial use in the GitHub repository at https://mtefagh.github.io/swiftcore/.