Dr. Maria Anisimova

Maria’s project focused on detecting positive selection in protein coding genes.  Her work made use of maximum likelihood (ML) methods based on codon substitution models accounting for heterogeneous selective pressure across sites.  She deducted that if ML estimates indicate presence of positive selection and the likelihood ratio test (LRT) is significant, Bayes prediction could be used to identify sites under positive selection.


    During ther project, she examined the accuracy and power of LRTs for positive selection and Bayes prediction of residues under positive selection.  She found that the use of X2 for significance testing makes the LRT conservative, especially for small samples of closely related lineages.  She found however that, if a large number of lineages of sufficient divergence are analyzed, the power of the LRT could be as high as 100%.  She also found that both accuracy and power of Bayes prediction are low for data containing only few similar sequences, but sampling a large number of lineages improves the performance substantially. She concluded that multiple models of heterogeneous selective pressure among sites should be applied in real data analysis and that ML models are phylogeny-based and do not incorporate recombination.  


To evaluate the effect of recombination on the LRTs and Bayes prediction, she simulated data using a coalescent model with recombination.  The LRT was found to be robust to low recombination rates.  However, she found that for higher rates, the type-I error rate can be very high.  Identification of sites under positive selection by the Bayes method was less affected by recombination than is the LRT.  Finally, she tested the hepatitis D antigen gene (HDAg) for positive selection.  Maria found sites predicted to evolve under positive selection in immunogenic domain and in the N-terminus region with reported antigenic activity.  At the end of the project, Maria concluded that no significant evidence of recombination was found.

Maria is currently at the Institute of Computational Science in Zurich, Switzerland. She is a member of the CBRG group of Prof Gaston Gonnet. She is interest in statistical genetics, evolutionary biology, molecular evolution and genomics. Her current research interests involve stochastic modelling of molecular evolution and statistical testing in phylogenetics. She is especially interested in application of likelihood and Bayesian techniques to these areas.

Maria's webpage: http://people.inf.ethz.ch/anmaria/index.html

Page last modified on 24 aug 09 11:05