Re: What practical experiments are biochemists involved with?

Hello James,

I could give you just a list of a bunch of methods (there is actually one 
by the end of this text) but I thought it could be interesting to develop 
your question a little. 

Biochemistry is the ”chemistry of life”. Strictly speaking it's about 
correlating chemical structure with biological activity at the molecular, 
sub-cellular, cellular and organism level. It's boundaries to other fields 
of research are being blurred more and more as we speak. Not that the 
boundaries have ever been tight. Molecular biology, cell biology, physical 
chemistry, organic chemistry, analytical chemistry, computer science, 
physiology, pharmacology etc. can be combined alone or together with 
biochemistry. This logic can also be turned around: biochemistry is one of 
the older chemical sciences addressing the chemistry of life and so could 
be said to have spawned some of the others... I won't mention any 
particular one as I don't want to get any hate mail.

There are maybe some methods that can still be considered ”pure” 
biochemistry: organelle or protein purifications. Structural analysis of 
proteins. Kinetic analysis of bindings and reactions. But even people 
concentrating on these, will be meddling with other methods too.

A great need for number-crunching has come with the advent of two major 
areas of research: modelling of structures and processes, coupled to the 
analysis of results from the big ”omics”.

Genomics was first: mankind had found a fast way to sequence DNA and went 
about trying to get a global picture of the genes: the genome. Scientists 
started with a medically important bacterium with a relatively small total 
amount of DNA (H. influenzae). It was sequenced in 1995. Many followed, 
including our dear ol' homo sapiens in 2001. To date there are some 150+ 
species sequenced. First number crunching is to put the bits of sequenced 
DNA into the correct order. This can be done ”manually” for a small number 
of DNA bases, but as the number of bits soars, it gets rapidly out of 
hand. Then you have to decide where a gene starts and ends within this 
sequence: you have to go looking for particular patterns, using for 
instance neural networks to find them. Next you might want to know which 
genes are turned on or turned off in particular situations, increasing or 
decreasing the number of proteins in a cell. Enter functional genomics, 
transcriptomics, proteomics and a load of other omics such as glycomics 
that studies the vast amount of sugar polymers and their role; 
metabolomics wants to chart the fate of all the small organic molecules in 
a cell. You even have bibliomics that applies the data-mining techniques 
from the other omics to extract new information from already published 
articles, a metascience that is also bound to gain in value.

When modelling molecules and molecular interactions, you are essentially 
trying to re-create a biological process inside a computer. In physical 
chemistry they have come pretty far at modelling a handful of such ”huge” 
entities as helium and lithium atoms. Imagine trying to do the same thing 
with a protein, composed of thousands of atoms of many different kinds and 
you soon realize that the project is daunting. This is the holy grail of 
protein modelling: to be able to calculate the 3D-structure of a protein 
just from it's chemical formula. We are still quite far from this goal. 
However, if one already has the structure of the protein from, for 
instance, crystallographical data, one can do simpler modellings (you'll 
still need something like a Cray supercomputer to do it, mind you) that 
would allow you to see how a particular protein interacts with its 
environment. The true structural data gives you a snap-shot of the 
protein, but you want to know how it moves or how it's active areas work. 
Big pharmaceutical companies do this to see how  new compounds might bind 
to a target protein, so that they can make a first selection of candidates 
before going through the hassle of actually doing a biochemical test. One 
could also use it to understand the inner workings of a protein. How it 
might work as a pump for instance. One also tries to model specific 
chemical processes inside a cell, some have even set as a goal to model a 
complete cell, with all it's metabolic processes. This would allow you to 
predict how changing the amounts of a particular molecule would affect the 
cell, either to mimic a disease, or how to inverse the effect of disease, 
or the metabolic effect of a pharmaceutical compound.

OK. So much for the long answer, maybe mainly for you guys who later 
stumble upon the question/answer, but which made James crazy. Here are some 
typical methods in biochemistry:

Electrophoresis (there are many kinds)
Centrifugations (separations based on weight and density)
PCR (polymerase chain reaction)
Immunoassays (use of antibodies to quantitate your experiments)
HPLC and other preparative separations
TLC (thin layer chromatography)
DNA synthesis
Protein synthesis
Cell culture
Microscopy, all kinds
Spectrophotometry
SPR (Surface plasmon resonance)
NMR (nuclear magnetic resonance), MS (mass spectrometry)
X-ray crystallography

I probably missed out on some very important applications, but maybe other 
scientists might fill the gaps. Last I'd like to direct you to a page I 
found hugely interesting while ”researching” for this little overview: 
www.genomicglossaries.com Here you will find a lot of information. Good 
luck with your report!

Kind regards,

Erik