CASCADE, microorganism, enrichment, biogas, animal manure, anaerobic digestion
Computer Modeling: We applied a machine learning system – Computer Assisted Strain Construction and Development Engineering, or CASCADE for short, to a collection of organisms and found correlations and predictive patterns between the organisms’ genetic information and their metabolic behavior. We compiled and gathered five types of genetic information of 366 organisms. We also compiled metabolic pathways from public databases, such as KEGG, and investigated metabolic reconstructions for the organisms with only genome information. We defined a measure called “average metabolic efficiency” (AME) for a given organism and found it highly correlates with the metabolic capabilities in real life. We built a series of predictive models using the five genetic information sources. Our predictive models can help biologists go beyond E. coli which is considered a golden standard, for example, modify E. coli more effectively by learning a desired behavior and their contributing factors across a collection of organisms. Compared to the traditional biological approaches, this approach is able to learn the behavior from relatively distant organisms other than their direct mutants. This might lead to better understanding of the actual underlying genetic mechanism. Applications of CASCADE to Biogas Production: Biogas production from anaerobic digestion is a traditional method to gain clean energy. Anaerobic digestion of animal manure has been extensively researched and demonstrated. However, there is lower gas production because of low biodegradability and unsteady bioprocess about manure digestion. Many methods, for example, pre-treating dairy manure by chemical or physical ways, are used to increase biogas production. A bioaugmentation method by adding bacteria selected by CASCADE to animal manure for biogas production is discussed in this paper. We use biological data that link the microorganisms to their capabilities of metabolism, e.g. digesting or producing substances of interest, and to perform an initial screening and selection organisms for applications that require these capabilities. Bacteria selected by CASCADE aimed to anaerobic digestion of cow manure are added to a continuous digestion process. The experiments showed that compared to the control of traditional anaerobic digestion, biogas production of added group increased by 40% and methane contents of biogas were nearly equal to the control experiments. Related DGGE analysis showed that addition of selected bacteria caused the difference of dominant microoragnisms during the digestion process which was useful to the anaerobic bacteria. The results showed that added bacteria selected by CASCADE is an effective way to increase biogas production from cow manure.