|
Form Determines Function in the World of Proteins
"Determining how to use the linear amino acid sequence of a protein to
predict its structure is the most significant problem in computational
biology," claims Robert Jones, director of computing at Darwin Molecular
Corporation. A key component in scientists' investigation of protein structure at
Darwin is Mathematica.
Researchers statistically analyze the sequence and known properties of a
protein to derive constraints on how far apart certain amino acids must be
in the structure. Using distance geometry methods, they work on finding a
three-dimensional configuration consistent with their matrix of distance
constraints.
"Mathematica is a ready toolkit of powerful mathematical
functions that makes it easy to try new techniques," says researcher Matt
Fitzgibbon. "The programming facilities allow us to rapidly prototype new
ideas and make changes to the algorithms. And the notebooks simplify the
development by recording our results for future analysis."
"Understanding protein structure helps us understand how proteins
function," explains Fitzgibbon. And that understanding gives scientists
insight that may lead to discoveries of new ways to fight viral infection, of methods of transporting therapeutics in the body, or of ways to curb
autoimmune disease.
Key features of Mathematica used:
- Numeric--matrix functions (particularly eigenvector and
eigenvalue), linear algebra functions
- Graphic--3D polylines to show derived structures
- Programming
- Notebook--"extremely useful for storing and annotating
intermediate steps in algorithm development"
|