engleski

Statistical evaluation of experimental data based on the high-resolution melting approach - application to BRCA1 polymorphisms

Mutations in BRCA1 (BReast CAncer Type 1 Susceptibility Gene) are responsible for approximately 40% of inherited breast cancers and more than 80% of inherited breast and ovarian cancers. We analyzed distribution of polymorphisms of BRCA1 gene among normal healthy population of Croatian women. For detecting polymorphisms we used a method based on high-resolution melting approach. Polymorphisms are distinguished from the wild- type genotype by changes in melting curve shape or position. Although every variant is confirmed by direct DNA sequencing, selecting PCR products for sequencing depends on visual ability of examiner to distinguish different melting patterns. This could be unreliable because sometimes melting profiles can be indiscernible. To verify our clustering (determination of how many different types of polymorphism do we have in each exon) and classification (determination of what type of polymorphism does certain DNA sample have) based on visual distinguishing melting patterns and direct DNA sequencing we used a few free open source data mining software (Weka, Tanagra...) and different statistical algorithms. Some of our results with application possibilities will be presented.

statistical analysis ; BRCA1 ; polymophism ; high-resolution melting

nije evidentirano