Re: culturing microorganismes
In GE, your major interest in growing microorganisms would probably be to multiply a target genetic sequence. If you're engineering bacteria, the easiest multiplying host is E. coli; if you're working on plants, try A. tumefaciens. (Don't even bother considering animal GE at this stage.) E. coli is far easier to work with than A. tumefaciens.
A. tumefaciens is used in plant biology only for integrating your gene of interest into the plant. Its not used for any subcloning work because it grows at a much slower rate, does not give high yields of plasmids, and on the whole is much more complicated.
Rather, all subcloning work is done in E. coli, even cloning the gene of interest into a suitable binary vector, which is then moved into Agrobacterium only for plant transformation.
DNA extraction is largely an analytical technique; you're better off buying your insertion genes. If you want to do something original and *have* to take your own from another organism, you'll need to extract it, but it's a very tricky procedure and you won't be able to tailor it at home anyway, so it'll be substandard. In general, extraction is used to confirm the presence of the gene in the organism (PCR is the simplest method), or for sequencing (you'll need to send it away for that).
I depends. First off, anyone who wants to try their hand at genetic engineering as a hobby or in their garage should stay away from unsequenced organisms. Its going to be hit or miss. Standard procedures like PCR are going to be complicated by the fact that you do not have reference genome to design you're primers to. Don't even think of trying to get some sequencing done. For traditional Sanger sequencing, you'll need primers. But since you don't have a reference genome, you'll have to use random primers which will give you random sequence that is essentially meaningless. Next gen sequencing could give you the entire genome, but would cost you a few thousand just to get the sequencing and then the amount of data alone requires servers dedicated solely to assembling. You're computer would crash trying to do it. Your other option would be to use the cloning techniques of the ancients, but I wouldn't recommend this for a beginner because they are so damn difficult.
In other words, the key is to play around with a sequenced organism, like E. coli. Yeast would be another excellent choice for beginners. If you want to do plants, then use Arabidopsis thaliana. Its fully sequenced and annotated to a very accurate degree and is easily to work with. All you need to genetically engineer A. thaliana is some Agrobacterium. I would not recommend mammals, genetic engineering is extremely difficult, requiring you to work with embryos, which is simply beyond the abilities of beginners.
My advice for the beginners would be to work with E. coli and yeast.
So this is where it depends.
If they are working with a sequenced organism, then they can easily clone a gene from the DNA. All they need is the ability to do PCR, restriction enzymes, a vector plasmid, and ligase.
If you buy an "insertion gene" usually these are cDNA clones in a standard vector, not designed for expressing the gene or any sort of genetic engineering other than amplification of the gene in E. coli.
To actually do anything with the gene, you need to clone it into a vector that will allow gene expression in the organism you wish to genetically engineer. So if you want to express in E. coli, then you need a vector that contains a promoter allowing for expression. You'll probably also want to tag it with a flourescent tag as well, like GFP.
Most "genes" that you can order are not in these vectors. So even if you ordered the gene, you would still have to clone it into another vector, which would require at the very least, restriction enzymes, a vector plasmid, and ligase. So if you have the capability to do PCR, then you might as well start from the beginning, its not going to be that much harder.