Have you ever wondered how to automate your biomolecular simulations in an efficient way so that you have more time for other aspects of your industrial project? Adam Hospital (AH), one of our BioExcel partners at the Institute for Research in Biomedicine in Barcelona, explains the capabilities of the bioBB and how to get started in the following interview:
What is the bioBB software library?
AH: The bioBB software library is a collection of Python wrappers on top of popular biomolecular simulation tools such as GROMACS for molecular dynamics (MD), pmx/GROMACS for free energy calculations or HADDOCK for protein-protein docking. The library offers a layer of interoperability between the wrapped tools, which make them compatible and easily interconnectable to build complex biomolecular workflows.
What are the advantages of bioBB-based workflows?
AH: Our workflows are reusable and reproducible, easy to build, and enable exascale computing. A workflow can be packed into a single Python script with a defined Conda environment or into a CWL (Common Workflow Language) specification file for wider sharing. Building the workflows is feasible via interactive GUIs such as Jupyter notebooks. A web server-based solution and a KNIME-based solution are being developed to make them more broadly accessible. Additionally, the workflows are compatible with HPC-ready workflow managers such as PyCOMPSs.
For which molecular modelling tasks do bioBB exist?
AH: So far, we have developed a number of different bioBB modules: for MD with GROMACS, for free energy calculations with pmx/GROMACS, for analyses using analysis tools of GROMACS and Ambertool’s cpptraj, to modify or extract information from a PDB, to prepare small molecules for MD simulations, as well as for checking and modelling protein structures. A bioBB on protein-protein docking with HADDOCK is being developed.
In which context have bioBB been particularly useful?
AH: We use the bioBB routinely in BioExcel workflows for different scientific use cases. For example, the effect of mutations on the structural flexibility of pyruvate kinase was studied, and features extracted from MD were used to predict the pathogenicity of mutations. More recently, we have been using bioBBs to predict drug resistance caused by mutations in the kinase domain of the Epidermal Growth Factor Receptor. Further details can be found here.
How easy is it to build such a workflow?
AH: It requires joining the different bioBB in a way that makes sense to a molecular modeller. You can find examples in the various demonstration workflow tutorials.
Where can I download the bioBB software library?
AH: Several solutions are provided for download: Source code at Github, Bioconda packages, Docker/Singularity/Biocontainers, and bioBB cloud. The bioBB cloud we are developing is the solution for combining one or several workflows with a GUI. The idea is to have a portable virtual machine that can easily be run in-house. The BioExcel industry user support would be happy to provide a customized solution based on such a virtual machine.
How do I start using bioBB?
AH: We have prepared several tutorials on the installation in different operating systems, demonstration workflows, and CWL, that can be useful for getting started. The demonstration workflows include a protein MD setup, automatic ligand parametrization, a protein-ligand complex MD setup, and free energy calculations for protein mutations. Virtual training on the use of the bioBBs is available. You can also add other tools that are most useful to your work to the building block library: feel free to have a look at our template on how to build your own building block.
Where can I get advice and support on bioBB?
AH: You can ask your technical questions via the Ask BioExcel Forum, which is a community support forum focusing on computational biomolecular research, particularly biomolecular simulation using high-performance computing. Just make sure not to post any confidential project information. If you would like to receive customized bioBB training, feel free to get in touch with firstname.lastname@example.org.
For further information:
- Andrio P et al. (2019): BioExcel Building Blocks, a software library for interoperable biomolecular simulation workflows. Sci Data. 10; 6(1): 169.