Predictive Patterns Among Microorganisms: Data Sciences for Screening Smart Bacteria for Methanogenesis and Wastewater Treatment

Abstract

“Smart” microorganisms are named for their extraordinary ability to generate energy and materials like electricity, hydrogen, methane, cleaning wastewater and proteins. Unlocking predictive patterns between microorganisms’ genetic fingerprints and their metabolism from compiled databases could provide revolutionary screening methods for discovering smart microorganisms. In this paper, we show a self-awareness concept and theory of natural swarm intelligence (SI) that can be used to discover authoritative and popular information as well as emerging and anomalous information when traditional connections among information nodes (e.g., hyperlinks or citations) are not available. The different categories of information can be all high-value depending on the application requirements. A self-awareness of swarm intelligence is a data-driven framework, modeled and measured using a recursive distributed infrastructure of machine learning. The combination of the machine learning and swarm intelligence are extended and enhanced in a completely new perspective. We built a data model from USPTO database, NCBI database, JGI (Joint Genomic Database) and KEGG database, as well as our own bio-database. We applied our big data biotechnology called CASCADE to microorganism populations using a measure we termed average metabolic efficiency (AME), which highly correlates with real life metabolic capabilities. We used the data models to select microbial consortia for wastewater treatment using the swarm intelligence of microbes. The collective behaviors of the selected microbes are used for cleaning wastewater and convert bio-wastes to usable energy. In methane experiments, we found that selected microbs are not only consistent with current scientific selection, but also allowed prediction for two additional microorganisms not previously selected. This technology can potentially identify mixtures of microorganisms that work more powerfully than single ones and dramatically speed up the discovery process.