The old way was to culture whatever earth you grabbed. The trouble is, most of it did not grow, although some did. When they perfected high speed sequencing the

The old way was to culture whatever earth you grabbed. The trouble is, most of it did not grow, although some did. When they perfected high speed sequencing they could then lyze out all the DNA from all the organisms living there and sequence and catalog them. They then take all the organisms they could grow from the same sample (representational sample), and then deduct all those sequences from the total. That was when they realized that they had only grown 10-15% of all the various organisms. There are hundreds of growth media, all painstakingly crafted by trial and error to grow particular organisms. Some organisms only grow by parasitism of others, some by eating others, and thus will only grow on a special media to fit this. It can take years to craft a growth medium. To this day we have many organisms that affect mankind that we can not grow in culture. The amassed sequence data can be analysed to see what sort of proteins they will grow. They can then try to make them - if they can - some need chaperone proteins that lead them into their correct folding pattern. No chaperone protein = no correct fold = wrong activity etc. It is very hard to 'uncook' the soup that made the amassed sequence data, but you may be able to identify promising proteins and tag those sequences in the living bacteria with a fluorescent dye that is taken up by the organism. Shine the right light on it and that organism will light up and you may be able to isolate it or hopefully, many of it. Repeat again and again gradually tease out these new organisms. With number to play with you can try to make them grow by analyzing their metabolism to see what might feed them and well fed ones will reproduce.