Non-culturable microorganism screening system based on generative adversarial network principle

Abstract

The invention discloses a non-culturable microorganism screening system based on a generative adversarial network principle, which performs data processing on non-culturable microorganisms to obtain metabolic network and genome data, inputs the metabolic network and genome data into a model shown in figure 1 to obtain a series of culture medium components, and performs automatic sampling, plate sealing, cultivation and detection to obtain a cultivation result. The results are input into a training network to improve a parameter optimization model. Data distribution difference exists between non-culturable microorganisms and culturable microorganisms, the system utilizes migration learning and based on the generative adversarial network principle, the non-culturable microorganisms are takenas a target domain data set, the culturable microorganism data set is taken as a source domain data set, the characteristic space distribution difference between a source domain and a target domain is reduced through a discriminator principle of the generative adversarial network, knowledge migration from the culturable microorganisms to the non-culturable microorganisms is completed, the automation system is enabled to be more complete, and the purpose of improving the accuracy of predicting a culture medium formula is achieved.

Description

technical field [0001] The present invention relates to technical fields such as bioinformatics, artificial intelligence, biological control technology, automatic control, etc., and specifically relates to a non-culturable microorganism screening system based on the principle of generative confrontation network. Background technique [0002] As we all know, various microorganisms exist in every corner of the world. It provides many conveniences to humans, such as penicillin, which is a metabolite of actinomycetes, can inhibit the growth of pathogenic bacteria, and lactic acid bacteria ferment carbohydrates into lactic acid, which is a good helper for making yogurt; but it also brings certain benefits to humans. Trouble, such as loss of breeding, water pollution, moldy food, etc. At present, less than 1% of all microorganisms can be cultivated in the laboratory, and very few can be cultivated industrially, which hinders people's understanding of the use of beneficial bacteri...

AI Technical Summary

Problems solved by technology

However, many microbial groups are very distantly related to laboratory strains and belong to brand-new families, orders, and phyla, which leads to large differences in the distribution of genome sequences between non-culturable microorganisms and laboratory strains, which will greatly reduce the accuracy of neural network predictions. Spend

Images