University of Twente Student Theses
Frequentist vs. empirical Bayes : a comparison of procedures for high-dimensional two-sample statistical inference, in particular for microRNA data
Sun, W.T. (2022) Frequentist vs. empirical Bayes : a comparison of procedures for high-dimensional two-sample statistical inference, in particular for microRNA data.
|
PDF
12MB |
| Abstract: | MicroRNAs are short strands of RNA, and nearly 2000 human microRNAs have been identified. As their levels in bodily fluids can change with the presence of cancer, they have potential to be used in non-invasive diagnosis. For this, datasets consisting of microRNA levels of healthy people and cancer patients need to be analyzed to classify which microRNAs correlate significantly with a disease---an example of a two-sample statistical inference problem. In this paper we compare two procedures to such problems: frequentist and empirical Bayes. We create simulated datasets similar to a microRNA dataset, analyze them using the two procedures, and compare the results. We find that the two procedures produce differing classifications. At the same significance level, the empirical Bayes procedure tends to classify more microRNAs as correlating, but can produce erroneous results if there are too few people per group; the frequentist procedure tends to classify microRNAs as non-correlating but is consistent. Applying the procedures on a lung cancer microRNA dataset, we find hsa.miR.10a.5p, hsa.miR.204.3p, hsa.miR.424.3p, and hsa.miR.509.3p as microRNA plausibly correlated with lung cancer. We conclude that although improvements are necessary, the empirical Bayes procedure has the potential to solve two-sample statistical inference problems and find correlating microRNAs well. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 31 mathematics |
| Programme: | Applied Mathematics BSc (56965) |
| Link to this item: | https://purl.utwente.nl/essays/92207 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page