engleski

Challenges of high-throughput glycomics - what to expect with large datasets

Understanding complex molecular mechanisms underlying common diseases requires an integrated analysis of large scale multiomics datasets. While genomics and proteomics have advanced rapidly, because of a limited set of tools, techniques and resources glycomics was for a long time lagging behind. Recent technological and methodological developments have enabled glycomics to join the other high-throughput omics. With a growing number of high-throughput profiling methods it is becoming increasingly evident that glycosylation analysis does not necessarily have to be time-consuming, labor-intensive, expensive and generally overwhelming. However, processing a large number of samples often reveals methodological weaknesses (like problems with rare events and outliers) which usually do not appear in small-scale studies. Therefore, generating high-quality glycomics data in a high-throughput fashion requires protocols dealing with the aforementioned problems while satisfying time constraints. Furthermore, combining different fields and their expert knowledge (computer science, statistics, chemistry, biology, etc.) with as much as possible automatization is of utmost importance. Based on our experience in dealing with large datasets, the most common problems and appropriate solutions will be discussed, including: proper study design (randomization and blocking), reproducible and robust sample preparation (experimental vs. biological variation), accurate glycan identification (data processing of UPLC/LC-MS data), quantification (automatic integration), and thorough quality control (normalization and batch correction). Identification and appropriate management of these critical steps has greatly improved the quality of our glycomics data and facilitated large-scale glycoprofiling.

glycomics; high-throughput; large scale studies

nije evidentirano