2023_Fall_Zubaer_L10_DNA_part1.pdf

Computational Molecular Microbiology (MBIO 4700) ABDULLAH ZUBAER UNIVERSITY OF MANITOBA Working with sequences Sequence databases Sequence comparison • Pairwise alignment • Multiple sequence alignment Phylogenetic tree Bayes’ theorem https://lavanya.ai/2019/05/16/bayes-theorem/ Bayesian meth...

Computational Molecular Microbiology (MBIO 4700) ABDULLAH ZUBAER UNIVERSITY OF MANITOBA Working with sequences Sequence databases Sequence comparison • Pairwise alignment • Multiple sequence alignment Phylogenetic tree Bayes’ theorem https://lavanya.ai/2019/05/16/bayes-theorem/ Bayesian method for phylogenetics ▪ For example, we have three species and an outgroup ▪ Here are all possible trees The Phylogenetic Handbook. by Fredrik Ronquist, Paul van der Mark and John P. Huelsenbeck, 2009. Bayesian method for phylogenetics ▪ The goal is to obtain the posterior probability distribution, which is the final result of the analysis based on the prior, the data and the model The Phylogenetic Handbook. by Fredrik Ronquist, Paul van der Mark and John P. Huelsenbeck, 2009. Bayesian method for phylogenetics ▪ Here is the posterior probability distribution ▪ However, for real instances of phylogenetic tree inference, it is actually impossible to calculate the posterior probability analytically The Phylogenetic Handbook. by Fredrik Ronquist, Paul van der Mark and John P. Huelsenbeck, 2009. MCMC method ▪ If we let the MCMC sampling run for a large enough number of iterations , the amount of time it spends sampling a particular parameter value is proportional to the posterior probability of that parameter value “Rerun” (generations) and sample trees along the way. Heuristic (trial and error) in approach – trees are continuously remodelled to achieve higher probabilities From: Analysis and Visualization of Tree Space Syst Biol. 2005;54(3):471-482. doi:10.1080/10635150590946961 Syst Biol | © 2005 Society of Systematic Biologists Figure 8 Progress in a Bayesian MCMC analysis. The progress in the search can be visualized in the Tree Set Visualization program, as a demonstration of how an MCMC analysis functions. In the visualization, the progress of the chain through treespace moves from regions of low optimality scores (blue) to regions of high optimality scores (red). “Keep the best and “burn” the rest: Summarize Probabilities and the tree topologies to get Consensus trees with “posterior probabilities (PP)” From: Analysis and Visualization of Tree Space Syst Biol. 2005;54(3):471-482. doi:10.1080/10635150590946961 Syst Biol | © 2005 Society of Systematic Biologists Figure 8 Progress in a Bayesian MCMC analysis. The progress in the search can be visualized in the Tree Set Visualization program, as a demonstration of how an MCMC analysis functions. In the visualization, the progress of the chain through treespace moves from regions of low optimality scores (blue) to regions of high optimality scores (red). PP PP and consensus trees MOLECULAR BIOLOGY AND EVOLUTION, VOLUME 26, ISSUE 10, OCTOBER 2009, PAGES 2299– 2315, HTTPS://DOI.ORG/10.1093/MOLBEV/MSP145 MrBayes http://mrbayes.sourceforge.net/ MrBayes Files have to be in NEXUS format (PAUP format) - AliView or DAMBE (has file converter) - MAFFT has an option to convert file program (command driven) - need manual Manual (can download from site – tutorial quite good) MrBayes – - Bayesian analysis (find the tree that fits the data) “Tree robot” evaluates tree topologies (tree space) until a set of trees are generated with overall “good” probabilities. - Computationally intensive – calculations done in parallel “chains”. (see NOTES on commands) MrBayes commands ▪ execute filename.nex ▪ prset aamodelpr=mixed ▪ mcmc ngen=20000 ▪ sump ▪ sumt BEAST – calculating evolutionary rates and estimating molecular clocks (emergences of disease outbreaks etc.) Good for estimating / calculating evolutionary rates. BEAUti & The BEAST Time scaled phylogenies “utility” needed to prepare input for BEAST https://github.com/beast-dev/beast-mcmc http://beast.community/ Checkpoint: Phylogeny With regards to phylogenetic analysis: Can you explain in one or two sentences? Maximum likelihood Distance methods Maximum parsimony methods Bayesian methods Bootstrapping Posterior probabilities