Shulou(Shulou.com)06/01 Report--

Editor to share with you how Iqtree2 uses the new model to build a phylogenetic tree efficiently. I believe most people don't know much about it, so share this article for your reference. I hope you can learn a lot after reading this article. Let's learn about it.

There are a variety of algorithms for phylogenetic inference (phylogenetic inference), from the simplest UPGMA method, to the adjacency method (neighbor joining), the maximum reduction method (maximum parsimony), to the complex maximum likelihood method (maximum likelihood) and Bayesian inference method (Bayesian inference). Each method has many optional implementation tools. Without exception, all of these methods follow a rule: the more accurate they are, the slower they are. With the analysis of more and more large data, it is necessary for users to find an algorithm that is as fast as possible and the error rate can be tolerated.

Bayesian method can infer the time tree accurately, but the speed is the slowest, while the adjacency method and the maximum parsimony method are prone to long branchattraction,LBA when dealing with sequences with great differences, so the maximum likelihood method is still the most commonly used method at present. At present, the tools to build trees by maximum likelihood method are MEGA, PhyML, RaxML and so on. However, the setting of these software parameters is very complex, especially the selection of nucleotide and amino acid substitution matrix often has a great impact on the accuracy of the results, but it is often difficult for ordinary users to choose. Moreover, the analysis speed of these tools is also slow. In recent years, a new phylogenetic inference tool Iqtree (http://www.iqtree.org/) has emerged quietly, which simplifies the parameter setting, can help users choose the best evolutionary model, and has a great improvement in speed.

The latest version is Iqtree2, which is installed as follows:

After the tar-zxvf iqtree-2.0.6-Linux.tar.gzcd iqtree-2.0.6-Linux is unzipped, the iqtree execution file is located in the bin/ directory. The usage is as follows: iqtree [- s ALIGNMENT] [- p PARTITION] [- m MODEL] [- t TREE].-s: sequence alignment files (multiple files are supported for comma separation, or folders containing comparison files) Optional PHYLIP, FASTA, NEXUS, CLUSTAL, MSF--seqtype: sequence type, optional BIN, DNA, AA, NT2AA, CODON, Morph defaults to automatic detection-o: outgroup list, comma separated between different species-- prefix: result file name prefix-- seed: random number seed, mainly for debugging purposes-- mem: maximum available memory in G, M or%-- redo: ignore check rewrite output file, default is off That is, from the last unexpected break-T: the number of cores used to run the program, you can set a specific number or AUTO (recommended). The default is 1--threads-max: the maximum number of cores that can be used. By default, it defaults to all cores-- fast: fast mode, similar to FastTree-b: nonparametric bootstrap, greater than or equal to 100Mub: ultra-fast bootstrap, greater than or equal to 1000--bnni: use NNI to optimize ultra-fast bootstrap trees. Match-B use-alrt:SH approximate likelihood ratio test repetition times-m: model selection, set MF to automatically select the best model without building a tree Set up MFP to automatically detect the best model and build a tree. In addition, you can set up specific models, or multiple optional models, such as-m LG,WAG--ancestral: reconstruct based on the ancestral state of empirical Bayes and then see how the tool is used. Let's start with the simplest building: iqtree-s example.phy-T AUTO

If you set up to automatically select the best model and build a tree:

Iqtree-s example.phy-m MFP-T AUTO

Select the best model and output only the model selection results:

Iqtree-s example.phy-m MF-T AUTO

Iqtree tests up to 546 protein models and gives the best model, and the results are as follows:

Use the bootstrap bootstrap method to calculate node support (similar to RaxML):

Iqtree-s example.phy-m MFP-b 100-T AUTO

SH approximate likelihood ratio test is used to calculate the support rate of nodes:

Iqtree-s example.phy-m MFP-- alrt 100T AUTO

Use two methods to calculate node support at the same time:

Iqtree-s example.phy-m MFP-- alrt 100b 100-T AUTO

Use the ultra-fast bootstrap bootstrap method to calculate the node support rate:

Iqtree-s example.phy-m MFP-B 1000-- bnni-T AUTO uses the above settings to build 120 tandem protein trees with 500 genomes in about two days. The ultrafast bootstrap1000 times of the ultra-fast bootstrap method is about 10 times faster than that of the ordinary bootstrap method, which is the unique algorithm of this software, so the ultra-fast bootstrap method of Iqtree is generally used to build the tree.

These are all the contents of the article "how Iqtree2 uses the new model to efficiently build a phylogenetic tree". Thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.