Enzymatic activity profiles are the result of many factors defined by both the enzyme and substrate in question. We recently collected a large amount of activity data for dozens of haloalkane dehalogenases--enzymes with great potential for biodegradation toxic compounds. The goal of this diploma thesis is to explore the potential of using machine learning for predicting activity profiles for new haloalkane dehalogenases based on the substrate molecular features and enzyme sequence-based features.

The graduate will solve the following specific tasks:
- study of problems and literature search in the field of enzyme engineering and machine learning,
- generation of features for enzymatic reactions for substrates and enzymes used in the experiments,
- training of machine-learning based predictors and analysis of their output and predictive power.

To solve the practical part of the diploma thesis, the student will use the Python environment and scikit-learn library.

The thesis will be written in English.

Recommended literature:
- Yang M, Fehl C, Lees KV, Lim EK, Offen WA, Davies GJ, Bowles DJ, Davidson MG, Roberts SJ, Davis BG. Functional and informatics analysis enables glycosyltransferase activity prediction. Nature chemical biology. 2018.
- Buryska T, Vasina M, Gielen F, Vanacek P, van Vliet L, Jezek J, Pilat Z, Zemanek P, Damborsky J, Hollfelder F, Prokop Z. Controlled Oil/Water Partitioning of Hydrophobic Substrates Extending the Bioanalytical Applications of Droplet-Based Microfluidics. Analytical chemistry. 2019.
- Marsland S. Machine learning: an algorithmic perspective. CRC Press; 2009.