Microbiome based identification, monitoring and enhancement of fermentation processes and products

A method comprising:

• a) receiving a set of samples comprising agricultural material;
• b) extracting nucleic acid material from each of the set of samples, with use of a homogenization process;
• c) barcoding said nucleic acid material with a barcoding process configured to correct sequencing errors and detect multi-bit errors and increase sequencing depth performance;
• d) amplifying said extracted and barcoded nucleic acid material to generate a 16S library and an internal transcribed spacer (ITS) library from said nucleic acid material;
• e) pooling the material of the 16S library with the material of the ITS library to generate a pooled library;
• f) sequencing the pooled library within a single run of a high-throughput sequencer, thereby obtaining a set of nucleic acid sequence reads of 16S and ITS genes of microorganisms represented in the set of samples;
• and
• g) applying the set of sequence reads representing microorganisms in each sample to a trained machine learning model, wherein the trained machine learning model comprises network architecture and regression architecture and is trained on 16S and ITS sequence data associated with soil status, agriculture product quality, contamination, and treatment response,
• the output of the trained machine learning model providing characterization of bacterial and fungal microorganism abundances for each of the set of samples.