Protein Language Models and Artificial Intelligence as Tools for Intracellular Protein Location Prediction in Complex Plastids

Nikhil THORAT

Diplomová práce

Abstract:

This thesis examines the application of artificial intelligence-based methods, which utilizes the protein sequences as input to predict the subcellular localization in complex plastids like diatoms. It also explores various traditional methodologies that use evolutionary information and then focuses on AI research so far in this specific domain. It combines cellular biology with artificial intelligence approaches such as protein language models using a limited but carefully curated dataset from an existing tool. The proposed approach aims to achieve robust predictive performance and perform a possible comparison with fixed decision boundary tool that has a hybrid approach of combining AI and rule-based approach.

Abstract:

This thesis examines the application of artificial intelligence-based methods, which utilizes the protein sequences as input to predict the subcellular localization in complex plastids like diatoms. It also explores various traditional methodologies that use evolutionary information and then focuses on AI research so far in this specific domain. It combines cellular biology with artificial intelligence approaches such as protein language models using a limited but carefully curated dataset from an existing tool. The proposed approach aims to achieve robust predictive performance and perform a possible comparison with fixed decision boundary tool that has a hybrid approach of combining AI and rule-based approach.

Keywords

Subcellular localization, Complex plastids, Diatoms, Plastid targeting, Periplastidic compartment (PPC), Signal peptides, Transit peptides, Protein sequence analysis, Protein language models (PLMs), ProtT5, ESM-1b, Transfer learning, Sequence-only prediction, Homology-aware cross-validation, GraphPart, N-terminal targeting signals, Machine learning in bioinformatics, Deep learning for protein sequences

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 31. 1. 2026

Obhajoba závěrečné práce

• Vedoucí: Ansgar Gruber, Dr. rer. nat.

JIHOČESKÁ UNIVERZITA V ČESKÝCH BUDĚJOVICÍCH

Přírodovědecká fakulta

Magisterský studijní program / obor:
Artificial Intelligence and Data Science / Artificial Intelligence and Data Science