Study-unit ARTIFICIAL INTELLIGENCE IN CHEMISTRY
| Course name | Chemical sciences |
|---|---|
| Study-unit Code | A003532 |
| Curriculum | Comune a tutti i curricula |
| Lecturer | Gabriele Cruciani |
| Lecturers |
|
| Hours |
|
| CFU | 6 |
| Course Regulation | Coorte 2023 |
| Supplied | 2024/25 |
| Learning activities | Affine/integrativa |
| Area | Attività formative affini o integrative |
| Sector | CHIM/06 |
| Type of study-unit | Opzionale (Optional) |
| Type of learning activities | Attività formativa monodisciplinare |
| Language of instruction | Italian |
| Contents | The main aim of the course is to provide a detailed overview of the use of artificial intelligence in chemistry and related subjects, showing how artificial intelligence can provide solutions to either simple or complex chemical problems. The current limits of artificial intelligence will also be discussed. In the laboratory part, software and platforms based on artificial intelligence will be used aimed at molecular design, the study of chemical and biological reactions and dissemination in the chemical field. General topics covered. Introduction on the history of artificial intelligence in chemistry. Most common approaches of artificial intelligence in chemistry. Data management. Structure-property relationships and predictive models. Artificial intelligence for molecular design. Artificial intelligence for chemical and biological reactions. Artificial intelligence in scientific dissemination. |
| Reference texts | Students can use the following texts to consult the topics discussed: H.M. Cartwright, Machine Learning in Chemistry. The impact of artificial intelligence, Royal Society of Chemistry. 2020 Artificial intelligence in drug discovery, ed. By Nathan Brown, Royal society of Chemistry, 2021. Teaching material of the course is provided directly by Lecturer in electronic form as pdf. |
| Educational objectives | The primary objective of the teaching is to provide students with the conceptual foundations to consciously use Artificial Intelligence tools for chemical applications and related subjects, understanding their potential and current limitations. At the end of the course the student should be able to understand the main AI strategies applied in chemistry, and know how to use software for molecular design, for the study of chemical reactions (enzymatic and non-enzymatic), the study of structure-property relationships. |
| Prerequisites | To follow and learn the contents of the course in a more convenient and profitable way, the student should possess the following basic knowledge: Knowledge of the fundamentals of chemistry; Knowledge of basic synthesis concepts in organic chemistry; Knowledge of computer chemistry elements. This knowledge is an important prerequisite for the student who intends to follow the course successfully. |
| Teaching methods | The Artificial Intelligence in Chemistry course includes laboratory exercises and is organized as follows: -Lectures in the classroom on all topics of the course lasting 2 hours each. -Laboratory experiences on the computer, at a single station or in groups of two, depending on the number of students. -Ongoing evaluation of the students' level of learning through collegial resolution of case studies. This activity is a form of exam training. |
| Other information | Gabriele Cruciani: gabriele.cruciani@unipg.it |
| Learning verification modality | The exam includes only an oral test, which consists of a discussion lasting approximately 30 minutes aimed at assessing the student's ability to describe and apply the concepts acquired. For the laboratory part, students will have to answer a questionnaire of 5 questions, three of which are multiple choice and two open, relating to the laboratory experiences carried out. This formula replaces the classic laboratory reports, and the answers will represent a starting point for discussion during the oral exam. |
| Extended program | Introduction to the course. History of artificial intelligence. History of artificial intelligence in chemistry. First approaches to artificial intelligence in chemistry. Definition of computational sciences. Pattern recognition concept. Description of molecules and macromolecules. Big data management. AI approaches for drug design. Prediction of non-covalent interactions. Protein-substrate interaction. Docking. Virtual screening. Structure-property and structure-activity relationships. Machine learning using PCA, LDA, PLS, PLS-DA, O-PLS methods. Neural networks and Deep Learning. AI strategies for chemical reactions. Description of a chemical reaction. How to formulate a retrosynthesis problem. Prediction of single-step or multi-step retrosynthesis. Reaction site prediction. AI to predict by-products or analyze impurities. AI strategies for predicting biotransformations AI in the prediction of exposure and reactivity factor, conformational factor and protonation state. Laboratory experiments The laboratory experiences are aimed at a deeper understanding of the concepts described in the theoretical part. Molecular design, prediction of molecular properties, prediction of retrosynthesis procedures and by-product analysis, prediction of metabolic transformations by enzymes. Use of AI tools for scientific presentations for educational purposes. |