|
 |
|
|
|
||
|
|
 |
 |
|
 |
|
|
|
|
|
||||||||||||||||||||||||||||||||||||
|
|
This method is an improving of Graziano Pesole's method [1]. It's based on cluster analysis (Complete Linkage algorithm), that requires a similarity matrix D containing distance between each pair of sequences of mRNA. It is a bottom-up method. It starts with n atomic cluster containing a single sequence of mRNA. For every step two clusters r and s such as D(r,s) = Max{di,j: where sequence r is in cluster r and sequence j is in cluster s} is minimum are merged in a new cluster. The process is stopped when diameter of new cluster obtained by every possible merging is greater than threshold (it's an input parameter). After cluster analysis, backtranslation is performed on sequences of homogeneous pool. In BBOCUS homogeneous pool is the cluster that has maximum diameter. Because it contains sequences aren't similar too much. Instead in Graziano Pesole's method the choose of homogeneous pool was made by a biologist. Codon Usage Table (CUT) is created through sequences of homogeneuos pool. CUT contains several fields: frequency, standard deviation, etc. Generally, to backtranslate an amino acid in a protein the codon chosen it is the one has maximum frequency. |
|||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||
| The backtranslation project has been realized by research group of Prof. Ferro, research group of Prof. Purrello, together two BS students Loredana Facciolà and Laura Patelmo. | ||||||||||||||||||||||||||||||||||||
 |
||||||||||||||||||||||||||||||||||||
