We develop and analyse neuroevolutionary systems to advance understanding of the capacity of evolutionary processes to generate increasingly intelligent behaviours in virtual creatures.
Our research into open-ended evolution looks to overcome the severe scaling problems exhibited by today's evolutionary algorithms, and toward evolving true artificial intelligence through natural selection. We developed the first closed artificial system to pass Mark Bedau et al's established statistical “ALife Test” for unbounded evolutionary dynamics . Earth's biosphere (through fossil-record databases) is the only other system to have passed the advanced form of this test. This is a very significant result: potentially a second example of unbounded evolution. The creation of a system capable of passing the test had been identified by Bedau, Snyder and Packard as “among the very highest priorities of the field of artificial life” . We made the test well-grounded even for long-term unbounded evolution in artificial systems, through the first ever method of computing individual genes' adaptive ('normalized') evolutionary activities [2003, 2006]. Our work in this area currently focuses on drawing generalized conclusions about open-ended evolution that were previously impossible given just the real-world example.
Our earliest work on the neuroevolution of articulated virtual creatures produced the first realistic co-adapted behaviours to have been evolved using general purpose neurons to control 3D physically simulated agents . The previous need for ad hoc (problem-specific) neurons was a barrier to the long-term evolution of general behaviours. Our research in this area is now focused on the incremental neuroevolution  of complex reactive and deliberative behaviour  coupled with motor control: evolving articulated virtual creatures capable of performing tasks that cannot be solved by reactive mechanisms alone and which would traditionally have their solutions formulated in terms of search-based planning . In related work, our biologically inspired evolutionary robotics research focuses on biological mechanisms of navigation .
The group also undertakes research into the evolution of social systems and social learning. We have presented the first example of behaviour inaccessible to incremental genetic neuroevolution alone being evolved through the addition of cultural transmission, in the form of learning by imitation ; the parallel use of transcription errors (noise in the genotype to phenotype map) was a key innovation here. And we have provided the first definitive answer to the previously open question of whether or not the “variability selection hypothesis” is sufficient to explain the adoption of social learning in increasingly variable environments . Our work on agent-based simulation of social systems has also focused on modelling the evolution of cooperation. In particular, we have studied the role of environment induced uncertainty and communication complexity in supporting the emergence of cooperation, and analysed factors that facilitate the emergence of cultural innovations.
We develop and analyse evolutionary systems to advance understanding of natural and artificial evolutionary processes.
Through EPSRC project EP/H031936/1 Evolution as an Information Dynamic System (2010-2013), we worked with partners Drs Chris Knight, Rok Krasovec, Roman Belavkin and (now) Professor John Aston, from the Universities of Manchester, Middlesex and Warwick (now Cambridge). Our work on mutation rate plasticity established optimal mutation rate control functions, for both artificial landscapes and natural landscapes defined by DNA-transcription factor affinities, and demonstrated general applicability to biology . The Evolutionary Systems Group used tables of DNA to protein binding affinities to evolve DNA sequences in computer simulations, and a meta-Genetic Algorithm  to evolve fitness-dependent mutation rate mappings that show a close match with the research team's predictions from information theory. We used Nvidia's CUDA many-core (GPGPU) technology to speed up computational experiments from 1.4 years (maximum, per run) to 3 days, enabling a step change in the rate at which the research was able to advance. The team began the follow-on BBSRC project BB/L009579/1 The theory and practice of evolvability: Effects and mechanisms of mutation rate plasticity in Febrauary 2014. In this project we are primarily investigating the effects and mechanisms of density-dependent mutation rate plasticity, although we first extended our previous work to determine optimal mutation rate control under selection .
Through the above Evolution as an Information Dynamic System project, the team also investigated the relationship between mutation rate, genetic loss and population size. Above a critical mutation rate (CMR) fitter genotypes may be outcompeted by those with greater mutational robustness. Expressed in terms of fitness landscapes, narrow high fitness peaks may be lost by a population while broader, lower peaks are maintained; this phenomena is referred to as “survival of the flattest”. We established that CMR has an exponential dependence on population size in haploid populations : that as population size falls, the CMR above which fitter alleles are lost transitions unexpectedly from near-constant (the previous assumption in evolutionary biology) to drop exponentially for small populations, leaving them spiralling toward extinction. We subsequently verified that our model closely reproduces the less significant but established mathematical relationship between population size and 'error threshold' (no mathematical model has yet been derived for CMR), and established that the result also holds for diploid populations . This and the team's related research has lead to our most recently awarded BBSRC project, BB/M021157/1 Adaptive landscapes of antibiotic resistance: population size and 'survival-of-the-flattest', which will start in August 2015. In this the focus is on the small population of a microbe with a newly arisen antibiotic resistance mutation.
We also use category theoretical models to study the evolution of complex systems. This work is inspired by the use of such approaches in type theory and system behaviour modelling. Our research on complex systems includes the network analysis of complex biological networks, for example protein interaction networks and ecological networks, and general research on formal modelling of the evolution of complex systems. Our network analysis research looks at the determination of structurally important components of biological networks in the context of drug discovery and environmental ecology. One of our earlier projects (eXSys) led to a spin-off company, e-Therapeutics Plc, which uses network analysis of proteomics data for discovery of new drugs and drug targets. More recently our network analysis research has focussed on the use of this approach to support the understanding of large-scale complex software systems by software engineers, for example in the context of matching revised requirements. We also use network data generated by dynamic analysis of large-scale software to evaluate and experimentally validate network analysis methods.
We recognise the importance of research-led teaching. Evolutionary Systems and Neural Networks are taught together in our undergraduate programmes, through Computational Intelligence modules embedded within our single and dual honours Computer Science programmes.
Externally Funded Projects
- Adaptive landscapes of antibiotic resistance: population size and 'survival-of-the-flattest', A Channon (Keele PI) and E Aston (Researcher Co-I) in collaboration with CG Knight and R Krasovec (Manchester) and R Belavkin (Middlesex). Funded by BBSRC, £621k, 2015-2018.
- The theory and practice of evolvability: Effects and mechanisms of mutation rate plasticity, A Channon and E Aston (PDRA 2014-2015) in collaboration with CG Knight and R Krasovec (Manchester) and R Belavkin (Middlesex). Funded by BBSRC, £465k, 2014-2017.
- Evolution as an Information Dynamic System, A Channon and E Aston (PhD student) in collaboration with R Belavkin (Middlesex), CG Knight (Manchester) and J Aston (Warwick, now Cambridge). Funded by EPSRC, £423k, 2010-2013.
- Protein interaction data visualization, P Andras. KTP project in collaboration with e-Therapeutics Plc, £60k, 2006.
- eXSys – Network analysis of protein interaction data, P Andras. Funded by EPSRC, £180k, 2003-2004.
- Network analysis of ecological systems, P Andras. Funded by DEFRA, £70k, 2003-2004.
Current PhD Students
- Mr Ben Jackson (supervisor: Dr Alastair Channon)
- Mr Ben Jolley (supervisor: Dr Alastair Channon)
Current Research Associate
- Dr Elizabeth Aston (supervisor: Dr Alastair Channon)
Past PhD Students
- Dr James Borg (supervisor: Dr Alastair Channon; graduated with PhD in 2018)
- Dr Adam Stanton (supervisor: Dr Alastair Channon; graduated with PhD in 2017)
- Dr Elizabeth Aston (supervisor: Dr Alastair Channon; graduated with PhD in 2014)
- Dr Thomas Miconi (supervisor: Dr Alastair Channon; University of Birmingham; graduated with PhD in 2008)
Past Research Associates
- Dr Iwo Bohr (supervisor: Prof Peter Andras; 2007; Newcastle University; currently postdoc at Newcastle University)
- Mr Neil Henderson (supervisor: Prof Peter Andras; 2006; Newcastle University)
- Mr William McElderry (supervisor: Prof Peter Andras; 2006; Newcastle University)
- Dr Greg Maniatopoulos (supervisor: Prof Peter Andras; 2005-2006; Newcastle University; currently postdoc at Newcastle University)
- Dr Olusola Idowu (supervisor: Prof Peter Andras; 2003-2005; Newcastle University)
- Dr Steven Lynden (supervisor: Prof Peter Andras; 2003-2004; Newcastle University; currently postdoc at National Institute of Advanced Industrial Science and Technology, Tsukuba, Japan)
- Dr Panos Periorellis (supervisor: Prof Peter Andras; 2003-2004; Newcastle University; currently program manager at Microsoft)
- Dr Agnes Madalinski (supervisor: Prof Peter Andras; 2003-2004; Newcastle University; currently postdoc at Otto-von-Guericke University of Magdeburg)
- Mr Robert Gwyther (supervisor: Prof Peter Andras; 2003-2004; Newcastle University; currently consultant at Accenture)
- Dr Chris Knight (University of Manchester)
- Dr Rok Krašovec (University of Manchester)
- Dr Roman Belavkin (Middlesex University)
- Professor John Aston (University of Cambridge)
- Professor Malcolm Young (e-Therapeutics Plc, Oxford)
- Dr Bruce Charlton (Newcastle University)
- Dr John Lazarus (Newcastle University)
- Dr Gilbert Roberts (Newcastle University)
- Stanton A, Harris LM, Graham G and Merrick CJ, Recombination events among virulence genes in malaria parasites are associated with G-quadruplex-forming DNA motifs, BMC Genomics 17:859, 2016. doi:10/bsnj
- Taylor T, Bedau MZ, Channon A, Ackley D, Banzhaf W, Beslon G, Dolson E, Froese T, Hickinbotham S, Ikegami T, McMullin B, Packard N, Rasmussen S, Virgo N, Agmon E, Clark E, McGregor S, Ofria C, Ropella G, Spector L, Stanley KO, Stanton A, Timperley C, Vostinar A and Wiser M, Open-Ended Evolution: Perspectives from the OEE Workshop in York, Artificial Life 22: 408-423, 2016. doi:10/bpqb
- Jolley BP, Borg JM and Channon A, Analysis of Social Learning Strategies when Discovering and Maintaining Behaviours Inaccessible to Incremental Genetic Evolution, in From Animals to Animats 14: Proceedings of the Fourteenth International Conference on Simulation of Adaptive Behavior (SAB 2016), pp. 293-304, Springer, 2016. doi:10/bpp9
- Stanton A and Channon A, Neuroevolution of Feedback Control for Object Manipulation by 3D Agents, in Proceedings of the Fifteenth International Conference on the Synthesis and Simulation of Living Systems (ALife XV), pp. 144-151, MIT Press, 2016. doi:10/bq8c
- Aston E, Channon A, Belavkin RV, Krasovec R and Knight CG, Critical Mutation Rate has an Exponential Dependence on Population Size for Eukaryotic-Length Genomes, in Proceedings of the Fifteenth International Conference on the Synthesis and Simulation of Living Systems (ALife XV), pp. 172-179, MIT Press, 2016. doi:10/bq8d
- Andras P, Social Learning, Environmental Adversity and the Evolution of Cooperation, in Proceedings of the Fifteenth International Conference on the Synthesis and Simulation of Living Systems (ALife XV), pp. 290-297, MIT Press, 2016. doi:10/bq8b
- Belavkin RV, Channon A, Aston E, Aston J, Krasovec R and Knight CG, Monotonicity of fitness landscapes and mutation rate control, Journal of Mathematical Biology 73(6): 1491-1524, 2016. doi:10/bd7r
- Stanton A and Channon A, Incremental Neuroevolution of Reactive and Deliberative 3D Agents, in Proceedings of the European Conference on Artificial Life 2015 (ECAL 2015), pp. 341-348, MIT Press, 2015. doi:10/59x
- Aston E, Channon A, Belavkin RV, Krasovec R and Knight CG, Optimal Mutation Rate Control under Selection in Hamming Spaces, in Proceedings of the European Conference on Artificial Life 2015 (ECAL 2015), pp. 640-647, MIT Press, 2015. doi:10/59z
- Andras P, Environmental Factors and the Emergence of Cultural-Technical Innovations, in Proceedings of the European Conference on Artificial Life 2015 (ECAL 2015), pp. 130-137, MIT Press, 2015. doi:10/bq79
- Krasovec R, Belavkin RV, Aston JAD, Channon A, Aston E, Rash BM, Kadirvel M, Forbes S and Knight CG, Mutation rate plasticity in rifampicin resistance depends on Escherichia coli cell-cell interactions, Nature Communications 5:3742, 2014. doi:10/skb
- Krasovec R, Belavkin RV, Aston JAD, Channon A, Aston E, Rash BM, Kadirvel M, Forbes S and Knight CG, Where antibiotic resistance mutations meet quorum-sensing, Microbial Cell 1(7): 250-252, 2014. doi:10/xzg
- Aston E, Channon A, Day C and Knight CG, Critical Mutation Rate Has an Exponential Dependence on Population Size in Haploid and Diploid Populations, PLOS ONE 8(12): e83438, 2013. doi:10/qqc
- Stanton A and Channon A, Heterogeneous complexification strategies robustly outperform homogeneous strategies for incremental evolution, in Advances in Artificial Life, ECAL 2013: Proceedings of the Twelfth European Conference on the Synthesis and Simulation of Living Systems, pp. 973-980, MIT Press, 2013. doi:10/nnb
- Andras P, Pakhira A, Moreno L and Marcus A, A measure to assess the behavior of method stereotypes in object-oriented software. Proceedings of the 4th International Workshop on Emerging Trends in Software Metrics (WETSoM 2013), pp.7-13, 2013.
- Borg JM and Channon A, Testing the Variability Selection Hypothesis: The Adoption of Social Learning in Increasingly Variable Environments, in Proceedings of the Thirteenth International Conference on the Simulation and Synthesis of Living Systems (ALife XIII), pp. 317-324, MIT Press, 2012. doi:10/m7p
- Kyriacou T, Using an Evolutionary Algorithm to Determine the Parameters of a Biologically Inspired Model of Head Direction Cells, Journal of Computational Neuroscience 32(2): 281-295, 2012. doi:10/bsmxfp
- Pakhira A and Andras P, Using network analysis metrics to discover functionally important methods in large-scale software systems. Proceedings of the 3rd International Workshop on Emerging Trends in Software Metrics (WETSoM 2012), pp.70-76, 2012.