In addition to a faculty winner and a nominee for the UvA Thesis PrizeExternal link, the Faculty of Science selects three other students for the Faculty of Science Master’s Thesis Prize each year. It is with great pleasure that we announce that Thomas Shehata (Forensic Science/Chemistry), Daniel Nuno Margato Guerra (Biological Sciences) and Oline Ranum (Artificial Intelligence) will be awarded this prize.

The faculty is proud of the amount of promising and original theses, and looks forward to review the upcoming crop of theses for the UvA Master’s Thesis Prize 2026. You can read more about Thomas’, Daniel’s and Oline’s theses below.

Oline Ranum - 3D Awareness, Geometry & Linguistics in Neural Sign Language Representations

Oline developed novel methods and datasets for sign language processing. Typical approaches use a 2D videos, her work builds and shows how to use small but high-precision datasets from 3D motion capture to improve sign language processing. To do so, she created novel machine learning pipelines and data analysis. This resulted in new open science datasets, as well as two publications. Her work helps build more inclusive natural language technologies.

Daniel Nuno Margato Guerra - Human-induced downsizing of animal communities disrupts plant frugivore trait matching in the tropics

In his thesis, Daniel studied how human impact affecting the prevalence of large-bodied animals affects the plants they feed on, by influencing the alignment between animal and plant traits relevant for seed dispersal of the plants. This is highly original research that Daniel expertly performed, and which resulted in a manuscript to be submitted by him to Ecology Letters, a top journal in his field. The work is not only of high scientific interest, but also very important from an applied perspective, given the high pressure of human action on large-bodied animal prevalence that, as Daniel showed in his work, may have a significant and unexpected effect on plants as well. The work was highly praised by both Daniel’s supervisors and marked with an impressive 9.0.

Thomas Shehata - An integrated Proteomic Workflow for Body Fluid Classification and Single Amino Acid Polymorphism Identification/Advancing towards Body Fluid Source Attribution

The report describes a highly innovative and original study to use state-of-the-art proteomics pipeline to detect the presence of human biological fluids and establish donorship also in case of mixed, contaminated stains. A very thorough study has been executed to establish the validity of the idea in a robust scientific manner. Detailed recommendations are provided to further develop and substantiate the suggested forensic proteomics approach.

The report is detailed, structured, very well written, and highly enjoyable to read (at least for a forensic scientist). It is also balanced in the use of figures, graphs, and tables, which are all of high quality. A substantial body of data is reported, analyzed, and interpreted. The report has formed the basis of a manuscript that recently has been submitted for publication in Forensic Science International: Genetics.

The student has completed a double MSc programme at the UvA, and in both Chemistry and Forensic Science he graduated with honour. For his Forensic Science MSc research project, he stayed in South Africa for half a year to develop forensic methods for investigating and fighting Wildlife Crime.