Seminar of the Laboratory on AI for Computational Biology
A seminar will be held at the Laboratory on AI for Computational Biology, at which laboratory employees - Burmak Karina and Shinkarev Elisei, will present their researches.
The seminar will take place on May 24 at 16:00.
Научно-учебная лаборатория искусственного интеллекта для вычислительной биологии: Стажер-исследователь
Title: Nanoplastic Identification With Mass Spectrometry
Speaker: Burmak Karina, Research Assistant at Laboratory on AI for Computational Biology
Annotation:
Currently, micro- and nanoplastics are a significant concern within the scientific community. Besides their ability to accumulate in human tissues and organs, they have the potential to cause serious diseases. Consequently, increasing attention is being paid to methods for detecting nanoplastics in human biological fluids. One such method is mass spectrometry, which allows for precise determination of the quantitative and qualitative composition of a sample, even at low concentrations. However, there are no computational algorithms and methods for the identification and annotation of nanoplastics. This seminar is devoted to a Python library which has been developed and implemented for the identification of polystyrene sulfonate (PSS) and perfluoroalkyl (PFAS) nanoplastics using mass spectrometry.
Title: "Deep learning for peptide retention time prediction"
Speaker: Shinkarev Elisei, Research Assistant at Laboratory on AI for Computational Biology
Annotation:
Current thesis work analyzes the application of deep learning for peptide retention time prediction and reviews the existing models: AutoRT and DeepLC. These models play a key role in improving the accuracy and reliability of peptide identification. They use deep learning techniques and training on large datasets with annotated peptide spectra, which allows them to capture complex patterns and features in the spectra, providing more accurate predictions of the feature - retention time. The work will involve setting up experiments on our dataset, examining the performance of the above models on peptides as well as their modifications, followed by comparison of distributions and analysis of results. How the predictions of a peptide and its modifications differ depending on the model used.