نبذة مختصرة : Introduction: In research using microarrays, various factors can lead to the introduction of systematic biases and errors in the study, so data quality control is an important step in any research. This removes incorrect or deficient data, such as false positive and negative values, which can significantly affect the results of the analysis, and by performing quality control in advance, we can ensure more reliable results of the analysis. The aim of this thesis is to prove the importance of data quality control for three different diseases obtained by microarrays for DNA genotyping of polymorphisms by comparing the results of association analysis before and after data quality control. Methods: For association analysis and data quality control the PLINK software and the R programming language were used. The results before and after quality control were graphically presented and the genomic inflation factor lambda was calculated. The difference between the results of association analysis was statistically evaluated using the Wilcoxon signed-rank test. Results: The results of association analysis obtained before and after data quality control differed statistically significantly (p-value for disease 1 < 2,2 × 10-16, p for disease 2 = 0,0003181 and p for disease 3 < 2,2 × 10-16). The difference can already be seen in the graphical display and from the lambda value, from which we can see that it is closer to 1 after quality control, which means that we successfully removed false positive and negative values. The results of the Wilcoxon signed-rank test confirmed that there is statistically significant difference between the results. Discussion and conclusion: In this thesis, we proved that quality control strongly influences the results of the analysis and is therefore a very important step in the analysis of data obtained with microarrays for genotyping DNA polymorphisms.
Uvod: V raziskavah z uporabo mikromrež lahko različni faktorji vodijo do uvedbe sistematičnih pristranskosti in napak v študiji, zato je pomemben korak v vsaki raziskavi kontrola kvalitete podatkov. Z njo se odstranijo napačni ali pomanjkljivi podatki, kot so lažno pozitivne in negativne vrednosti, ki lahko pomembno vplivajo na rezultate analize. S predhodno izvedbo kontrole kvalitete lahko zagotovimo zanesljivejše končne rezultate analiz. V naši nalogi smo želeli dokazati pomembnost kontrole kvalitete pri podatkih za tri različne bolezni, ki smo jih pridobili z mikromrežami za genotipizacijo DNA polimorfizmov, tako da smo primerjali rezultate asociacijskih analiz pred opravljeno kontrolo podatkov in po njej. Metode: Za asociacijsko analizo in kontrolo kvalitete podatkov smo uporabili program PLINK in programski jezik R. Rezultate pred kontrolo kvalitete in po njej smo grafično predstavili ter izračunali genomski inflacijski faktor lambda. Razliko med rezultati asociacijskih analiz smo statistično ovrednotili z Wilcoxonovim testom predznačenih rangov. Rezultati: Z Wilcoxonovim testom predznačenih rangov smo ugotovili, da se rezultati asociacijskih analiz, pridobljeni pred kontrolo kvalitete podatkov in po njej statistično značilno razlikujejo (p-vrednost za bolezen 1 < 2,2 × 10-16, p za bolezen 2 = 0,0003181 in p za bolezen 3 < 2,2 × 10-16). Razliko med podatki asociacijskih analiz, pridobljenimi pred kontrolo podatkov in po njej, smo dokazali tudi z grafičnim prikazom lambda vrednosti, iz katerega je razvidno, da je lambda vrednost po kontroli kvalitete bližje 1, kar pomeni, da smo uspešno odstranili lažno pozitivne in negativne vrednosti. Razprava in sklep: V nalogi smo dokazali, da kontrola kvalitete močno vpliva na končne rezultate analize in je zato zelo pomemben korak v analizi podatkov, pridobljenih z mikromrežami za genotipizacijo DNA polimorfizmov.
Processing Request
No Comments.