This is our BioNavigator
Enzymaster has developed a proprietary in-house computational toolbox for structure guided combinatorial library design and enhanced diversity generation.
The BioNavigator® is a set of state‑of‑the‑art methods for computational enzyme identification, in silico activity and selectivity prediction and stability calculation under process conditions. This includes bioinformatics driven hot spot identification, focused combinatorial library design and in silico loop design.
BioNavigator® is a set of state‑of‑the‑art methods for computational enzyme identification.
AI and structure‑based recombination algorithms are applied for in silico pre-screening to enhance the combinatorial sequence diversity coverage beyond traditional random and site-saturation mutagenesis approaches, thereby increasing the sequence space coverage and identifying all important sequence areas.
This approach allows the design of focused libraries (typically 1000–2000 variants/round) which reduce the number of required experiments and enrich the libraries with active hits, ensuring a maximum screening efficiency utilizing our BioEngine® technology. After each round of evolution, all confirmed hits are sequenced for knowledge generation and computationally analyzed by docking and (QM/MM)-MD Simulations on our in-house Linux cluster, as well as on the π-2-supercomputer at Shanghai Jiao Tong University or cloud based HPC, to analyze and improve our prediction and prescreening methods using our BioNavigator® technology.
Computer aided Evolution
Traditional random and site-saturation mutagenesis methods mostly cover only a small range of the possible combinatorial sequence space, often leading to a local, rather than the global maximum. With computer aided evolution, in silico pre-screening combined with careful computational analysis of all reaction and evolution data of each screening and evolution round improves the in silico model after each round. This builds up additional knowledge-based and artificial intelligence driven models, which speed up the evolution progress by optimized library design, thus reducing the required amount of directed evolution rounds and increasing the overall sequence space coverage.