Quantitative biology group, School of Biological Sciences, University of Liverpool

The quantitative biology group meetings are intended to provide a forum for:

-presenting results in more mathematical and statistical detail than is possible in the existing research groups
-developing the depth of mathematical knowledge in the School
-discussing possible modelling approaches for biological problems
-getting feedback on grant applications involving modelling and statistics
-developing ideas for improving teaching of mathematics and statistics in the School
-introducing new software that can be used for biological modelling

Autumn 2009

We have now merged the quantitative biology group meetings with the bioinformatics and systems biology meetings. We meet on alternate Tuesdays at 11 a.m. in Seminar Room 6, Life Sciences Building, to discuss topics anywhere on the interface between biology, mathematics, and computer science. For a current list of seminars, see the Liverpool bioinformatics website.

Spring 2009

Friday 27 March, 11 a.m., SR16 (third floor, Life Sciences). Anna Hickman (Earth and Ocean Sciences). A stochastic approach to modelling phytoplankton distributions in the ocean. Abstract: Despite the thousands of phytoplankton species that live in the world's oceans, most models incorporate only a handful of phytoplankton types. Here I will describe a novel approach being developed at MIT, in which thousands of phytoplankton ecotypes can be initiated. The characteristics for each phytoplankton ecotype are assigned stochastically via a series of coin-flips and random selection from plausible parameter ranges. I will show an application for an open-ocean environment and illustrate how the different types interact and compete with one another in the model environment. I will briefly discuss how the model helps us better understand phytoplankton populations as well as highlights what we are still to understand about phytoplankton physiology.

Friday 24 April, 11 a.m., SR2 (second floor, Life Sciences). Double bill (the whole session should be done within an hour).
1. Prudence Wong (Computer Science). Using AI techniques for siRNA design and motif finding.. Abstract: Short interfering RNAs (siRNAs) can be used to suppress gene expression and possess many potential applications in therapy, but how to design an effective siRNA is still not clear. Based on the MPI (Max-Planck-Institute) basic principles, a number of siRNA design tools have been developed recently. The set of candidates reported by these tools is usually large and often contains ineffective siRNAs. In view of this, we initiate the study of filtering ineffective siRNAs. The algorithm we designed is based on some new observations on the secondary structure, which we have verified by AI techniques (decision trees and support vector machines). We have tested our algorithm together with the MPI principles and the existing tools. The results show that our filtering algorithm is effective. Recently, we have applied AI techniques for motif finding, in particular, in the context of calcium binding sites. Many calcium binding proteins contain the DxDxDG motif. However the DxDxDG motif alone is not enough to confer the ability to bind calcium, it has been suggested that it may be possible to predict this ability using the sequence surrounding the motif. We again used AI techniques to analyze the data in an attempt to identify related features. Some preliminary results will be discussed.
2. Ozgur Akman (Centre for Systems Biology, University of Edinburgh). Oscillatory gene networks: insights from mathematical models. Abstract: Circadian oscillations are universal, controlling 24-hour rhythms of metabolism, physiology and behaviour in organisms ranging from humans to cyanobacteria. The regulatory gene networks underlying these oscillations have proved to be useful tools for quantifying the relationship between the structures of biochemical systems and their functional properties. By constructing mathematical models of key circadian organisms and analysing the models using techniques derived from nonlinear dynamics, insights can be obtained into the reasons why circadian networks have considerably more complex architectures than the minimal negative feedback loop required for entrainable, autonomous oscillations. Recent results obtained from modelling the fungus N. crassa and the plant A. thaliana support theoretical studies proposing that one of the possible benefits of the high feedback loop complexity observed in clock networks is the increased functional flexibility that such architectures confer. Moreover, the computational techniques developed to construct the models have potential applicability to a broader range of signalling pathways, particularly with respect to parameter-fitting and sensitivity analysis.
Details of his research interests are available at http://csbe.bio.ed.ac.uk/akman.php.

Meetings are co-organized by Matt Spencer and David Montagnes. Suggestions or volunteers for future meetings are welcome.

Past seminars: Autumn 2008

Friday 10 October, 2 p.m, SR3 [CHANGE OF TIME AND LOCATION]. Andy Jones (Pre-clinical Veterinary Science). Bioinformatics methods for annotating genomes using proteomics data. Slides. Abstract: The major bottleneck in many genome projects is the determination of accurate gene models i.e. finding the correct start codon, stop codon and intron-exon boundaries for every gene. Even genome sequences that have undergone significant manual curation still contain a large number of inaccurate gene models. Gene prediction algorithms continue to improve, but in the absence of experimental corroboration, hypothetical gene models are not highly reliable. In this seminar, I will demonstrate how protein expression data derived from proteome studies can be used to improve genome annotation, in so-called proteogenomic approaches. To date, several proteogenomic studies have been performed on various organisms, confirming the correct prediction of large numbers of genes. However, in a typical proteome study mass spectrometry (MS) data is usually searched against a database of predicted proteins. As such, if a gene sequence has not been predicted, or been incorrectly predicted, corresponding peptides will not be identified in a database search, limiting the utility of proteomics to inform the improvement of gene model predictions. We are developing bioinformatics methods that allow MS data to play an active role in gene model prediction. Our approaches can provide evidence that alternative gene predictions appear to be correct and can find previously unpredicted ORFs, demonstrated on studies of apicomplexan parasites (such as Toxoplasma gondii). I will present a summary of the research challenges and the future directions of our research.

2. Friday 24 October, 1 p.m, SR1. Harry Noyes (Biological Sciences). Modelling haplotype evolution. Abstract: Genetic recombination has traditionally be analysed in terms of centiMorgans. The advent of genome sequences means that it is now possible to study recombination on a physical rather than genetic map of the chromosome. I have written a Perl script that simulates the recombination process between arbitrary numbers of founders. A linear model of the evolution of mean length of haplotype derived from each parent was discovered that accurately predicted the simulated mean for up to about 50 generations. Since crossovers are Poisson distributed and haplotype lengths are exponentially distributed the mean could be used to predict the distribution  of both these parameters. The simulation was run for 10,000 generations to discover the evolution of haplotype length since the emergence of Homo sapiens. The development of a model for haplotype evolution over this period of time will be discussed.

Friday 7 November, 1 p.m., SR1. David Montagnes (Biological Sciences). Variable assimilation efficiency in predator-prey models. Abstract: Most predator-prey models assume that predators assimilate ingested prey with constant efficiency. We describe alternative empirical models in which assimilation efficiency may vary with prey density, and show that they are sometimes a better fit to measurements of ingestion and growth. We show how our models relate to the conventional approach, and discuss the consequences for population dynamics.

Friday 21 November, 1 p.m., SR1. Graeme Thorn and Rachel Bearon (Mathematical Sciences). Phytoplankton in turbulence. Abstract: Many harmful algal bloom-forming phytoplankton are motile: the interaction of the swimming with fluid flow can affect the long-term transport of individuals in non-trivial ways. This talk will discuss how motility interacts with a range of flows from simple shear to homogeneous turbulence, using experiments on the swimming of the multicellular alga Volvox in prescribed flows as an example system.

Friday 5 December, 1 p.m., SR1. Bakhti Vasiev (Mathematical Sciences). Pattern Formation in Excitable Systems with Applications to Developmental Biology. Abstract: Excitable media are characterized by homogeneous steady state, where over-threshold disturbances can result to formation of propagating waves or localized structures. Both types of patterns are important for mathematical studies in developmental biology. I show that propagating waves play critical role in development of social amoebae while localized structures govern processes during embryogenesis.

Friday 19 December, 1 p.m., SR1. Mike Begon (Biological Sciences). Metapopulation perspectives on the dynamics of plague: new explorations. Abstract: In ecology generally, 'metapopulation dynamics' has taken on an increasingly wide range of meanings. This seems to be reflected in the study of plague dynamics, where a number of groups have (apparently independently) developed a metapopulation approach. I briefly review these studies, with the aim of clarifying the different meanings that have been given by plague ecologists to the term 'metapopulation'. I then describe some new theoretical explorations that acknowledge both the metapopulation structure of many plague systems and that many (perhaps most) plague systems comprise more than one important host species. Particular attention will be paid to 'Allee effects' (inverse density dependence at low densities), both in the colonisation of empty sub-populations and in the transmission dynamics of plague itself, and to the important effects these may have on plague dynamics, including the introduction of multiple stable states and an increased sensitivity to the effects of second (and further additional) host species.

Past seminars: Summer 2008

Friday 9 May, 11 a.m. (different time), SR2. Ajanthah Sangaralingam (School of Biological Sciences). Why do models of gene gain and loss give the incorrect phylogeny?

POSTPONED (due to the maths not being ready): Andy Fenton (School of Biological Sciences). Title and date TBA.

Friday 30 May, 1 p.m, SR2. Simon Whelan (University of Manchester). Heterogeneity in sequence evolution.

Friday 6 June, 1 p.m, SR2. Jon Read (Veterinary Pathology). Networks and Epidemiology: trying to model human social mixing.

Friday 13 June, 11 a.m. (different time), SR2. Russell Hyde (Biomedical Sciences). Local and global structure in protein interaction networks.

Tuesday 17 June, 2 p.m. (different time), SR1 (different place). Melissa Penny (Swiss Tropical Institute). Modelling within-host dynamics of Plasmodium falciparum malaria infections (within a framework for the stochastic simulation of malaria epidemiology and control).

Friday 20 June, 1 p.m., SR1 (different place). Mike Speed (School of Biological Sciences). Quick optimisation of complex traits: some examples from anti-predator defence.

Past seminars: Autumn 2007

3 October 2007. Pawel Paszek. Delay induced asynchrony of NF-kappaB oscillations.

17 October 2007. Andy Fenton. Determining the optimal immune response against co-infecting parasites.

31 October 2007. Matt Spencer. Fitting stochastic predator-prey models to time-series data using sequential Monte Carlo.

14 November 2007. Caroline Horton. Modelling NF-kappaB signalling dynamics in response to differently timed stimuli.

28 November 2007. Geoff Parker. Using the Evolutionarily Stable Strategy approach with continuous strategies.

All at 11 a.m. in Seminar Room 2, Life Sciences Building. Everyone is welcome.

Home