Mathematical Models of Oversaturated Protein Spots

Abstract

Two-dimensional electrophoresis is used for protein separation in a special gel according isoelectric point and molecular mass, and remains one of most progressive technologies for protein analysis. As a result, gels are digitized and their images are process with software. The dynamic diapason of intensities of protein spots is very wide, thus software tools become too sensitive to various artefacts inevitable present in gel images. Often matching with reference or modelling of protein spot intensity distribution is done willing to get better protein spot estimates. Nevertheless that many protein spot models exist, they take into account only particular spot shape artefact. In the article modern protein spot models applicable in two-dimensional electrophoresis and DNR arrays are analysed. It is shown that symmetry is preserved only for protein spots in DNR arrays. Three new protein spots shape models – 4G, 4S and Bell shape – are presented in order to cope with shape distortion and asymmetry. Results of experimentation with 84 distorted protein spots applying three developed shape models and one of most popular – Flat-Top shape model are presented. It is shown that Bell shape model fits best to oversaturated protein spots while 4G shape model fits best to non-saturated spots and requires smallest computational time. Ill. 5, bibl. 11 (in English; summaries in English, Russian and Lithuanian).