Cipres Web Portal

---

RAxML Black Box

RAxML is a hybrid code created by a collaboration between Alexandros Stamatakis and Wayne Pfeiffer. The Science Gateway provides a familiar Black Box interface (at phylobench.vital-it.ch) and an advanced interface. Depending upon the job, this interface submits to Trestles, a large NSF XSEDE Resource. Trestles allows your job to run for up to 168 hours (7 days). The job is run using a hybrid mpi/pthreads RAxML, so run speed is 3 - 4.5 times faster than the old CIPRES cluster. RAxML on CIPRES Science Gateway ONLY accepts matrices in relaxed Phylip format. Tips for running on XSEDE here.

RAxML (Randomized Axelerated Maximum Likelihood) is a program for Maximum Likelihood-based inference of large phylogenetic trees. The program is explicitly being developed to efficiently infer trees for extremely large datasets, either in terms of the number of taxa and/or the sequence length. For example, a 25,000-taxon alignment of protobacteria with an alignment length of 1,500 base pairs had a run time on a single CPU of the cluster of only 13.5 days, with a memory consumption of only 1.5GB. Much additional information can be found on Alexis' RAxML page.

Through OpenMP-based parallelization of RAxML, the program can also efficiently exploit the SMP (Symmetric Multi Processing) capabilities of the cluster with its 8-way SMP nodes. This type of parallelism is especially useful for very long alignments. For example, together with Olaf Bininda-Emonds and Usman Roshan, we are currently working on a multi-gene analysis of almost 70 genes for a total of 2,100 mammal species.

Rapid bootstrap heuristics for RAxML are more than an order of magnitude faster than current algorithms. Computational experiments on 22 DNA and AA (amino acid) containing 125 up to 7764 sequences; the RBS inferences are between 8 and 20 times faster (average 14.73) than SBS analyses with RAxML and between 18 and 495 times faster than BS analyses with competing programs, such as PHYML or GARLI The performance improvement increases with alignment size.

RAxML includes support for DNA, Protein, Binary, and RNA Secondary Structures.

For RAxML 7.2+, selecting the GTRGAMMA model has a very different effect (command line = -m GTRGAMMA -x -f a). This option causes GTRGAMMA to be used both during the rapid bootstrapping AND inference of the best tree. The result is that it takes much longer to produce results using GTRGAMMA in RAxML 7.0.4, and the analysis is different from the one run using RAxML 7.0.4, where GTRCAT was used to conduct the bootstrapping phase. If you wish to run the same analysis you ran using RAxML 7.0.4, you must instead choose the model GTRCAT (-m GTRCAT -x -f a). In RAxML 7.2.+, this option is identical to choosing GTRGAMMA in RAxML 7.0.4 and below. In other words, the GTRCAT switch causes GTRCAT to be used during the rapid bootstrapping, but the program then switches to GTRGAMMA for the ML search.

Also, in RAxML 7.2.+, the -N autoWC option bootstopping is no longer available.  Instead, there are several options for using majority rule to set the threshold for bootstopping. RAxML 7.2.+ also allows the user the option of uploading a set of bootstrapped trees for a posteriori bootstopping using frequency and majority rule criteria. 

There is currently no manual for RAxML 7.2.+, but we have attempted top provide adequate Advanced Help section on the Science Gateway interface, with assistance from Alexandros Stamatakis. The manual for version 7.0.4 is available here: [pdf]

The RAxML home page is here. The RAxML Users Google Group is here:

INPUT = dna, protein, or mixed matrices in relaxed Phylip format.

Test input file (nucleic acid): raxml_input.phy

Optional Input: Partition File: a text file in the following format:
DNA, gene1=1-921
DNA, gene2=922-4015
(Note that "DNA" must be in all caps).

Test output file1 (run summary): RAxML_info_result.txt

Test output file2 (bootstrap result): RAxML_bootstrap_result.txt

Test output file3 (besttree): raxml_outtree.tre

If you use RAXML, please cite:

• Alexandros Stamatakis: ``RAxML-VI-HPC: Maximum Likelihood-based Phylogenetic Analyses with Thousands of Taxa and Mixed Models”. In Bioinformatics 22(21):2688-2690, 2006. (pdf)
• A. Stamatakis, P. Hoover, J. Rougemont. (2008) A Fast Bootstrapping Algorithm for the RAxML Web-Servers, Systematic Biology 57(5): 758-771.

Known Issues:

• When specifying more than one outgroup, do not introduce blank spaces. Use outgroup1,outgroup2 not outgroup1, outgroup2. In the latter case you will see the following error message "Error, you must specify a model of substitution with the '-m' option"
  
• There is a fundamental difference in the management of jobs after RAXML 7.1.0, and the nomenclature can be confusing for users familiar with 7.0.4 and below.
  Please see the RAxML page.

If there is a tool or a feature you need, let us know.