Glycobiology is dogged by the relative scarcity of synthetic, defined oligosaccha-9 rides. Enzyme-catalysed glycosylation using glycoside hydrolases is feasible, but is 10 hampered by the innate hydrolytic activity of these enzymes. Protein engineering 11 methods are applicable, though usually require prior structural knowledge of the 12 target enzyme and the use of powerful computing methods, and/or relies on extensive 13 screening methodologies. Here we describe a straightforward strategy that involves 14 rapid in silico analysis of protein sequences. The method pinpoints a small number 15 (<10) of mutant candidates aimed at diminishing hydrolysis and thus tipping the 16 reaction balance toward transglycosylation. Requiring no other significant prior 17 knowledge of the target enzyme than its sequence, the results reveal that the method 18 is quite generic, allowing the improvement of glycoside hydrolases that act on dif-19 ferent α-/β-pyranosides or furanosides. Moreover, the presented data support that 20 mutational hotspots that are validated in one enzyme can be transposed to other 21 related enzymes. 22 23 Glycosides are ubiquitous and abundant in Nature, being essential for biological interactions 24 and processes. Nevertheless, progress in glycobiology is hampered by the lack of synthetic 25 carbohydrates, an issue related to their complexity. Carbohydrates are composed of polyhy-26 droxylated units that exist in different forms (e.g. pyranoside or furanoside), interlinked in 27 a variety of regioselectivities, with the anomeric centres displaying either α-or β-anomeric 28 configurations 1. Faced with this high degree of complexity organic chemistry has developed 29 numerous glycosylation methodologies 2,3. These involve several synthetic steps, including 30 protection-deprotection cycles, are characterised by relatively poor overall yields, and gen-31 erate considerable amounts of waste. This is in stark contrast to polynucleotides and poly-32 peptides, both of which are accessible via automated chemical synthesis processes and 33 through in vivo biological synthesis. Unfortunately, unlike these biopolymers, most carbo-34 hydrates cannot be obtained using straightforward, generic technologies 4 amenable to auto-35 mation. 36 Enzyme-catalysed glycosylation offers an alternative to chemical methods. The nat-37 ural choice for this are glycosyltransferases (GTs) that are well-represented in a variety of 38 families in the CAZy database (http://www.cazy.org/) 5 , with each family potentially har-39 bouring numerous specificities 6. Nevertheless, GTs have proven to be rather difficult to han-40 dle in vitro and often require expensive nucleotide-glycoside donors 7. Therefore, glycoside 41