Secondary Structure Classification of Isoform Protein Markers in Oncology - Mathematical Approaches to Polymer Sequence Analysis and Related Problems - page 65

Tabl e 3. 5 Folding characteristics of the protein given in EMBL-EBI and isoforms
entries in the PDB databases according to the CASTOR algorithm
Isoform BRCA1 BRCA2 BRCA3 BRCA4 BRCA5

Length 1,860 1,860 760 1,846 720

BRCA1 - Identity % - 99.9 40:6 98:9 38:6

- Similarity % 99.9 40:6 98:9 38:6

-Gaps% 0.0 59:4 1:1 61:4

- Differences residues N 1 - 20 -

BRCA2 - Identity % - - 40:6 98:9 37:7

- Similarity % - 40:6 98:9 37:7

-Gaps% - 59:4 1:1 62:3

- Differences residues N - - 20 -

BRCA3 - Identity % - - 39:9 94:7

- Similarity % - - 39:9 94:7

-Gaps% - - 60:1 5:3

- Differences residues N - - - 40

BRCA4 - Identity % - - - 37:7

- Similarity % - - - 37:7

-Gaps% - - - 62:3

- Differences residues N - - - -

Folding segments
Helix folds 167 164 71 147 78

Strands folds 82 83 38 72 41

Other folds 199 198 69 177 73

in many cases these factors may be multi-determined, and the complete analysis of
these aspects should be studied. There may be some important dependencies be-
tween different factors, so the effect of a modification may be undetermined, and
this will give rise to random-like effects, as reported in many studies.
A more complete analysis and a precise formalization of the various factors that
may affect malignancies should be pursued. This will require a detailed study of
the phenomenon so that the predictions of the outcome can be determined with a
given precision, and therefore, there will arise a certain confidence in the role of
each aspect.
To carry out these types of studies, the utilization of the prediction of the sec-
ondary structure of the proteins should be considered as a useful stage toward the
recourse to tertiary and quaternary structures and a more precise formulation of the
factors to be examined.

3.6 Conclusions

A classification process called CASTOR is described. It is based on a nonlinear
binary optimization algorithm as an implementation of a model which is derived
on a strict statistical methodology avoiding extraneous assumptions, which could