Software created through or connected with the 2020 Science programme…

As described in our research manifesto, a novel computational approach is needed for doing "2020 science" - and this does not exist today. However, the groups involved in the programme have been - and continue to be - developing tools to facilitate novel approaches in computational bioscience. A selection of these tools are described on these pages.


HemeLB is a software package for modelling and simulation of haemodynamics in sparse geometries. To date, it has been mainly applied to the simulation of complex cerebrovascular networks and cerebrovascular malformations such as aneurysms. HemeLB includes: a) a GUI for simulation definition, b) a highly-optimised parallel lattice-Boltzmann (LB) solver including different LB collision operators, boundary conditions, and rheology models, c) a visualisation client allowing real-time visualisation and simulation steering, and d) a number of postprocessing tools. The package is currently developed at the Centre for Computational Sciene, UCL. 


Chaste (Cancer, Heart and Soft Tissue Environment) is a general purpose simulation package aimed at multi-scale, computationally demanding problems arising in biology and physiology. Current functionality includes tissue and cell level electrophysiology, discrete tissue modelling, and soft tissue modelling. The package is being developed by a team mainly based in the Computational Biology Group at the Department of Computer Science, University of Oxford, and development draws on expertise from software engineering, high performance computing, mathematical modelling and scientific computing.

Bacterial motility analysis tool

We have developed an efficient, open source implementation of a novel automated, non-parametric method for characterising the swimming patterns of flagellated bacteria.


Distribution Modeller (codename- HYKL)

Complex Species Distribution Modelling analyses is made accessible, via the browser, by the HYKL software. HYKL streamlines the tasks which require considerable engineering, allowing the user to concentrate on the science. You can build models using the user interface, or switch to the code view so that you can create custom models.


MEMOIR is a homology modelling pipeline designed for membrane proteins. The inputs are the sequence which is to be modelled, and the 3D structure of a template membrane protein. MEMOIR integrates our membrane protein software programs iMembrane (membrane annotation), MP-T (alignment), MEDELLER (core co-ordinate generation) and PyFREAD (loop modelling). Each of its constituent components is specifically developed for membrane proteins and outperforms non-specialised software. The MEMOIR pipeline has been described in the 2013 Web Server Issue of Nucleic Acids Research.


PyFREAD is a tool to model the structural loops in three-dimensional protein structures. It is a fast implementation of the fragment-based loop modelling algorithm FREAD. Given a loop's amino acid sequence and anchor co-ordinates (the co-ordinates of the residues on either side of the loop) FREAD can model loops at an accuracy of approximately 1 Angstrom.

iBatsID (Europe)

iBatsID (Europe) is a classification tool which uses ensembles of artificial neural networks (eANN's) to classify time-expanded recordings of bat echolocation calls from 34 European bat species.


MEDELLER is an algorithm for the co-ordinate generation step in the three-dimensional modelling of membrane protein structure. It takes a sequence alignment between a target and a template membrane protein and produces a 3D model of the target. Its strength lies in how it decides which template residues are likely to be structurally conserved. This is done by using information about the template protein's position within the lipid bilayer, predicted using iMembrane. This is combined with a set of environment-specific substitution tables and a rule-based algorithm described in the MEDELLER paper. Once the core model has been built, loops are added using the FREAD algorithm. Finally, for convenience, any remaining gaps are filled using Modeller.


iMembrane is a programme to predict the orientation of a given membrane protein within the lipid bilayer. This is useful in itself for biologists wishing to visualise a particular membrane protein in the context of the membrane. In addition, computational scientists can use the data generated by iMembrane to inform methods specific to membrane proteins (such as MEDELLER).

Quantifying Uncertainty

A cloud-computing implementation of a class of parallel Monte-Carlo Markov Chain methods called Generalized Metropolis-Hastings (GMH). Due to its parallel nature, GMH can be used for Bayesian estimation of the parameters of computationally-intensive biophysical models. The project also provides a biologically-plausible model of fly photoreceptor cells as an application. Computations can run on multi-core clusters in the cloud and are accessed via a web interface.