About this Event
MCB Seminar | Artificial Intelligence for Mass Spectrometry Proteomics Data Analysis
The Molecular and Cellular Biology Seminar series features Ibi M. Max-Harry discussing "Artificial Intelligence for Mass Spectrometry Proteomics Data Analysis" on Oct. 19 from 4:35 to 5:55 p.m.
Abstract: Over the years, there has been an increase in the use of Mass Spectrometry (MS) as a high throughput method for identifying proteins present in a wide range of biological samples. This has been particularly useful in the identification of biomarkers associated with certain diseases. However, the large volume of data produced by MS poses a data analysis problem for biologists. Most bioinformatic techniques used to analyze MS proteomic data involve the use of a protein sequence database software to identify the protein based on its peptide fragment mass. The downside to this method is that it can be inaccurate due to the high possibility of numerous proteins with similar sequence. Furthermore, although the detailed information about the exact protein is useful to understand biological pathways, it can be time consuming and unnecessary if the goal is to simply group samples based on their spectra profiling. Machine Learning involves training algorithms till they can make accurate predictions or decisions regarding inputted data. Deep neural networks (DNNs), a subset of machine learning are known for their higher accuracy compared to other machine learning algorithms due to the presence of multiple hidden layers in the network. Wang et al developed a DNN-based open-source method called MSpectraAI for classification of samples into various groups based on their MS spectra profile. They showed the accuracy of this method by using datasets of six tumor types (a total of 7,997,805 mass spectra). In future, MSpectraAI can be used for disease screening as well as the development of personalized treatment approaches to illnesses.
Wang, S., Zhu, H., Zhou, H. et al. MSpectraAI: a powerful platform for deciphering proteome profiling of multi-tumor mass spectrometry data by using deep neural networks. BMC Bioinformatics 21, 439 (2020). https://doi.org/10.1186/s12859-020-03783-0 Liebal UW, Phan ANT, Sudhakar M, Raman K, Blank LM. Machine Learning Applications for Mass Spectrometry-Based Metabolomics. Metabolites. 2020;10(6):243. Published 2020 Jun 13. doi:10.3390/metabo10060243
0 people are interested in this event