RAxML-HPC v.8 on XSEDE 8.2.12 Phylogenetic tree inference using maximum likelihood/rapid bootstrapping run on XSEDE Alexandros Stamatakis Stamatakis A. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models.Bioinformatics. 2006 Nov 1;22(21):2688-90 Phylogeny / Alignment http://icwww.epfl.ch/~stamatak/index-Dateien/countManual7.0.0.php raxmlhpc8_rest_xsede raxmlhpc_hybridlogicdhge10 perl $nchar < 500000 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "raxmlHPC-HYBRID_8.2.12_expanse" 0 raxmlhpc_hybridlogicdpge10 perl $nchar > 499999 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "raxmlHPC-PTHREADS_8.2.12_expanse" 0 raxmlhpc_hybridlogicphge10 perl $nchar < 200000 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "raxmlHPC-HYBRID_8.2.12_expanse" 0 raxmlhpc_hybridlogicppge10 perl $nchar > 199999 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "raxmlHPC-PTHREADS_8.2.12_expanse" 0 raxmlhpc_hybridlogicallplt10 perl $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "raxmlHPC-PTHREADS_8.2.12_expanse" 0 raxmlhpc_hybridlogicdhge10_scheduler1 scheduler.conf perl $nchar < 10000 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "jobtype=mpi\\n" . "mpi_processes=10\\n" . "threads_per_process=4\\n" . "cpus-per-task=4\\n" . "mem=77G\\n" . "node_exclusive=0\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicdhge10_scheduler2 scheduler.conf perl $nchar > 9999 && $nchar < 40000 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "jobtype=mpi\\n" . "mpi_processes=5\\n" . "threads_per_process=25\\n" . "cpus-per-task=25\\n" . "mem=243G\\n" . "node_exclusive=1\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicdhge10_scheduler3 scheduler.conf perl $nchar > 39999 && $nchar < 500000 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "jobtype=mpi\\n" . "mpi_processes=2\\n" . "threads_per_process=64\\n" . "cpus-per-task=64\\n" . "mem=243G\\n" . "node_exclusive=1\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicdpge10_scheduler scheduler.conf perl $nchar > 499999 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "threads_per_process=128\\n" . "cpus-per-task=128\\n" . "node_exclusive=1\\n" . "mem=243G\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicdplt10_scheduler1 scheduler.conf perl $nchar < 4000 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=8\\n" . "cpus-per-task=8\\n" . "node_exclusive=0\\n" . "mem=15G\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicdplt10_scheduler2 scheduler.conf perl $nchar > 3999 && $nchar < 16000 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=16\\n" . "cpus-per-task=16\\n" . "node_exclusive=0\\n" . "mem=30G\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicdplt10_scheduler3 scheduler.conf perl $nchar > 15999 && $nchar < 60000 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=24\\n" . "cpus-per-task=24\\n" . "node_exclusive=0\\n" . "mem=46G\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicdplt10_scheduler4 scheduler.conf perl $nchar > 59999 && $nchar < 300000 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=48\\n" . "cpus-per-task=48\\n" . "node_exclusive=0\\n" . "mem=92G\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicdplt10_scheduler scheduler.conf perl $nchar > 299999 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=128\\n" . "cpus-per-task=128\\n" . "node_exclusive=1\\n" . "mem=243G\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogicphge10_scheduler1 scheduler.conf perl $nchar < 3000 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "jobtype=mpi\\n" . "mpi_processes=10\\n" . "threads_per_process=4\\n" . "cpus-per-task=4\\n" . "mem=77G\\n" . "node_exclusive=0\\n" . "nodes=1\\n" raxmlhpc_hybridlogicphge10_scheduler2 scheduler.conf perl $nchar > 2999 && $nchar < 12000 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "jobtype=mpi\\n" . "mpi_processes=10\\n" . "threads_per_process=12\\n" . "cpus-per-task=12\\n" . "mem=243G\\n" . "node_exclusive=1\\n" . "nodes=1\\n" raxmlhpc_hybridlogicphge10_scheduler3 scheduler.conf perl $nchar > 11999 && $nchar < 30000 && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "jobtype=mpi\\n" . "mpi_processes=5\\n" . "threads_per_process=25\\n" . "cpus-per-task=25\\n" . "mem=243G\\n" . "node_exclusive=1\\n" . "nodes=1\\n" raxmlhpc_hybridlogicphge10_scheduler4 scheduler.conf perl $nchar > 29999 && $nchar < 200000 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "jobtype=mpi\\n" . "mpi_processes=2\\n" . "threads_per_process=64\\n" . "cpus-per-task=64\\n" . "mem=243G\\n" . "node_exclusive=1\\n" . "nodes=1\\n" raxmlhpc_hybridlogicppge10_scheduler scheduler.conf perl $nchar > 199999 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) perl "threads_per_process=128\\n" . "cpus-per-task=128\\n" . "mem=243G\\n" . "node_exclusive=1\\n" . "nodes=1\\n" raxmlhpc_hybridlogicpplt10_scheduler1 scheduler.conf perl $nchar < 3000 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=12\\n" . "cpus-per-task=12\\n" . "node_exclusive=0\\n" . "mem=23G\\n" . "nodes=1\\n" raxmlhpc_hybridlogicpplt10_scheduler2 scheduler.conf perl $nchar > 2999 && $nchar < 4500 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=24\\n" . "cpus-per-task=24\\n" . "node_exclusive=0\\n" . "mem=46G\\n" . "nodes=1\\n" raxmlhpc_hybridlogicpplt10_scheduler3 scheduler.conf perl $nchar > 4499 && $nchar < 15000 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=32\\n" . "cpus-per-task=32\\n" . "node_exclusive=0\\n" . "mem=61G\\n" . "nodes=1\\n" raxmlhpc_hybridlogicpplt10_scheduler4 scheduler.conf perl $nchar > 14999 && $nchar < 100000 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=64\\n" . "cpus-per-task=64\\n" . "node_exclusive=0\\n" . "mem=123G\n" . "nodes=1\\n" raxmlhpc_hybridlogicpplt10_scheduler scheduler.conf perl $nchar > 99999 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" perl "threads_per_process=128\\n" . "cpus-per-task=128\\n" . "node_exclusive=1\\n" . "mem=243G\\n" . "nodes=1\\n" infile Sequences File (relaxed phylip format) (-s) 1 infile.txt infile_regular Sequences File (relaxed phylip format) (-z) perl $select_analysis ne "J" perl "-s infile.txt" 1 infile_Joption Collection of trees file (-z) perl $select_analysis eq "J" perl " -z infile.txt" 1 runtime 1 scheduler.conf Maximum Hours to Run (click here for help setting this correctly) 0.25 Maximum Hours to Run must be less than 168 perl $runtime > 168.0 Maximum Hours to Run must be greater than 0.1 perl $runtime < 0.1 perl "runhours=$value\\n" The job will run on 40 processors as configured. If it runs for the entire configured time, it will consume 40 x $runtime cpu hours perl $nchar < 10000 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) The job will run on 125 processors as configured. If it runs for the entire configured time, it will consume 125 x $runtime cpu hours perl $nchar > 9999 && $nchar < 40000 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) The job will run on 128 processors as configured. If it runs for the entire configured time, it will consume 128 x $runtime cpu hours perl $nchar > 39999 && $nchar < 500000 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) The job will run on 128 processors as configured. If it runs for the entire configured time, it will consume 128 x $runtime cpu hours perl $nchar > 499999 && $datatype ne "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) The job will run on 8 processors as configured. If it runs for the entire configured time, it will consume 8 x $runtime cpu hours perl $nchar < 4000 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 16 processors as configured. If it runs for the entire configured time, it will consume 16 x $runtime cpu hours perl $nchar > 3999 && $nchar < 16000 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 24 processors as configured. If it runs for the entire configured time, it will consume 24 x $runtime cpu hours perl $nchar > 15999 && $nchar < 60000 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 48 processors as configured. If it runs for the entire configured time, it will consume 48 x $runtime cpu hours perl $nchar > 59999 && $nchar < 300000 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 128 processors as configured. If it runs for the entire configured time, it will consume 128 x $runtime cpu hours perl $nchar > 299999 && $datatype ne "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 40 processors as configured. If it runs for the entire configured time, it will consume 40 x $runtime cpu hours perl $nchar < 3000 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) The job will run on 120 processors as configured. If it runs for the entire configured time, it will consume 120 x $runtime cpu hours perl $nchar > 2999 && $nchar < 12000 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) The job will run on 125 processors as configured. If it runs for the entire configured time, it will consume 125 x $runtime cpu hours perl $nchar > 11999 && $nchar < 30000 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) The job will run on 128 processors as configured. If it runs for the entire configured time, it will consume 128 x $runtime cpu hours perl $nchar > 29999 && $nchar < 200000 && $datatype eq "protein" && ($choose_bootstop eq "bootstop"|| $bootstrap_value > 9 || $altrun_number > 9) The job will run on 128 processors as configured. If it runs for the entire configured time, it will consume 128 x $runtime cpu hours perl $nchar > 199999 && $datatype eq "protein" && ($choose_bootstop eq "bootstop" || $bootstrap_value > 9 || $altrun_number > 9) The job will run on 12 processors as configured. If it runs for the entire configured time, it will consume 12 x $runtime cpu hours perl $nchar < 3000 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 24 processors as configured. If it runs for the entire configured time, it will consume 24 x $runtime cpu hours perl $nchar > 2999 && $nchar < 4500 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 32 processors as configured. If it runs for the entire configured time, it will consume 32 x $runtime cpu hours perl $nchar > 4499 && $nchar < 15000 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 64 processors as configured. If it runs for the entire configured time, it will consume 64 x $runtime cpu hours perl $nchar > 14999 && $nchar < 100000 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" The job will run on 128 processors as configured. If it runs for the entire configured time, it will consume 128 x $runtime cpu hours perl $nchar > 99999 && $datatype eq "protein" && $bootstrap_value < 10 && $altrun_number < 10 && $choose_bootstop ne "bootstop" Estimate the maximum time your job will need to run. We recommend testing initially with a time less than 0.5hr test run because Jobs set for 0.5 h or less depedendably run immediately in the "debug" queue. Once you are sure the configuration is correct, you then increase the time. The reason is that jobs > 0.5 h are submitted to the "normal" queue, where jobs configured for 1 or a few hours times may run sooner than jobs configured for the full 168 hours. datatype Please select the Data Type protein dna rna binary multi dna 2 outsuffix Set a name for output files (-n) perl "-n $value" result Please enter a name for the output files (-n) perl !defined $outsuffix 1 nchar How many patterns in your data set? 1000 Please enter the number of patterns in your data set perl !defined $nchar Knowing the number of characters in your dataset helps us determine the most efficient way to run raxml. The number of patterns is the number of unique columns in the multiple sequence alignment matrix. You can get this number from the output of the intermediate results once a job begins. Entering the number of characters per taxon in your matirx, or 1000 as the number of patterns is an ok start. Look at the intermediate results, and see if that is reasonably close. If it is not, kill the job, and adjust the number. 1 MLsearch_CAT Enable ML searches under CAT (-F) perl $choose_bootstrap ne "x" && $choose_bootstrap ne "b" perl ($value)? " -F " : "" 0 This option allows one to enable tree searches under CAT for very large trees, as this saves memory. This option can also be used under GAMMA models to avoid thorough optimization of the best scoring ML tree at the end of the run. outgroup Outgroup (-o) ()one or more comma-separated outgroups, see comment for syntax) perl (defined $value)? " -o $value " : "" 10 The correct syntax for the box is outgroup1,outgroup2,outgroupn. If white space is introduced (e.g. outgroup1, outgroup2, outgroupn) the program will fail with the message "Error, you must specify a model of substitution with the '-m' option" number_cats Specify the number of distinct rate categories (-c) perl (defined $value)? " -c $value" : "" 25 2 perl ($datatype eq "dna" && $dna_gtrcat eq "GTRCAT") || ($datatype eq "protein" && $prot_sub_model eq "PROTCAT") || ($datatype eq "binary" && $bin_model eq "BINCAT") This option allows you to specify the number of distinct rate categories, into which the individually optimized rates for each individual site are thrown under -m GTRCAT. The default of -c 25 works fine in most practical cases. disable_ratehet Disable Rate Heterogeneity (-V) perl ($value)? " -V " : "" 0 2 perl ($datatype eq "dna" && $dna_gtrcat eq "GTRCAT") || ($datatype eq "protein" && $prot_sub_model eq "PROTCAT") || ($datatype eq "binary" && $bin_model eq "BINCAT") This option allows you to disable rate heterogeneity anong the sites. Valid for CAT model only. treetop Supply a tree (Not available when doing rapid bootstrapping, -x) (-t) perl $choose_bootstrap ne "x" perl " -t tree.tre" 2 tree.tre Specifies a user starting tree file in Newick format. Not available when doing rapid bootstrapping. Branch lengths of that tree will be ignored. Note that you can also specify a non-comprehensive (not containing all taxa in the alignment) starting tree now. This might be useful if newly aligned/sequenced taxa have been added to your alignment. Initially, taxa will be added to the tree using the MP criterion. The comprehensive tree will then be optimized under ML. provide_parsimony_seed Specify a random seed value for parsimony inferences (-p) perl $select_analysis ne "fe" && $select_analysis ne "fA" 1 Please provide a parsimony seed (-p) perl $altrun_number && !defined $parsimony_seed_val Please provide a parsimony seed (-p) perl $startingtreeonly && !defined $parsimony_seed_val Sorry, you cannot specify a starting tree (via the -t option above) and a random seed value (via -p) with the -f x option perl defined $treetop && $provide_parsimony_seed && $select_analysis eq "fx" Specify a random number seed. The -p option allows you and others to reproduce your results (reproducible/verifiable experiments) and will help Alexis debug the program. Do not use this option if you want to generate multiple different starting trees. parsimony_seed_val Enter a random seed value for parsimony inferences (-p "value" gives reproducible results from random starting tree) perl $provide_parsimony_seed perl ($value) ? " -p $value" : "" 12345 2 Please enter a random seed for the -p option (eg 12345) perl $provide_parsimony_seed && !defined $parsimony_seed_val rearrangement_yes Specify an initial rearrangement setting (-i) 0 number_rearrange Specify the distance from original pruning point (-i) perl (defined $value)? " -i $value" : "" 10 perl $rearrangement_yes Please specify the distance from original pruning point (-i) (default would be 10). perl $rearrangement_yes && !defined $number_rearrange 2 This option allows you to specify an initial rearrangement setting for the initial phase of the search algorithm. If you specify e.g. -i 10; the pruned subtrees will be inserted up to a distance of 10 nodes away from their original pruning point. If you dont specify -i; a "good" initial rearrangement setting will automatically be determined by RAxML. constraint perl !defined $binary_backbone && $select_analysis ne "y" Constraint (-g) constraint.tre perl defined $value ? " -g constraint.tre" : "" 2 This option allows you to specify an incomplete or comprehensive multifurcating constraint tree for the RAxML search in NEWICK format. Initially, multifurcations are resolved randomly. If the tree is incomplete (does not contain all taxa) the remaining taxa are added by using the MP criterion. Once a comprehensive (containing all taxa) bifurcating tree is computed, it is further optimized under ML respecting the given constraints. Important: If you specify a non-comprehensive constraint, e.g., a constraint tree that does not contain all taxa, RAxML will assume that any taxa that not found in the constraint topology are unconstrained, i.e., these taxa can be placed in any part of the tree. As an example consider an alignment with 10 taxa: Loach, Chicken, Human, Cow, Mouse, Whale, Seal, Carp, Rat, Frog. If, for example you would like Loach, Chicken, Human, Cow to be monophyletic you would specify the constraint tree as follows: ((Loach, Chicken, Human, Cow),(Mouse, Whale, Seal, Carp, Rat, Frog)); Moreover, if you would like Loach, Chicken, Human, Cow to be monophyletic and in addition Human, Cow to be monophyletic within that clade you could specify: ((Loach, Chicken, (Human, Cow)),(Mouse, Whale, Seal, Carp, Rat, Frog)); If you specify an incomplete constraint: ((Loach, Chicken, Human, Cow),(Mouse, Whale, Seal, Carp)); the two groups Loach, Chicken, Human, Cow and Mouse, Whale, Seal, Carp will be monophyletic, while Rat and Frog can end up anywhere in the tree. binary_backbone perl !defined $constraint Binary Backbone (-r) binary_backbone.tre perl (defined $value) ? " -r binary_backbone.tre" : "" 2 This option allows you to pass a binary/bifurcating constraint/backbone tree in NEWICK format to RAxML. Note that using this option only makes sense if this tree contains fewer taxa than the input alignment. The remaining taxa will initially be added by using the MP criterion. Once a comprehensive tree with all taxa has been obtained it will be optimized under ML respecting the restrictions of the constraint tree. partition Use a mixed/partitioned model? (-q) perl " -q part.txt" 2 part.txt This parameter allows you to upload a file that specifies the regions of your alignment for which an individual model of nucleotide substitution should be estimated. This will typically be used to infer trees for long (in terms of base pairs) multi-gene alignments. If DNA and protein mixed models are used together (for example) you should choose a model option based on the model of rate heterogeneity you want to use. If you specify either -m GTRCAT or PROTCAT, the CAT model will be used, if you specify -m GTRGAMMA or -m BINGAMMA, the GAMMA model will be used .... For example, if -m GTRGAMMA is used, individual alpha-shape parameters, GTR-rates, and empirical base frequencies will be estimated and optimized for each partition. Since Raxml can now handles mixed Amino Acid and DNA alignments, you must specify the data type in the partition file, before the partition name. For DNA, this means you have to add DNA to each line in the partition. For AA data you must specify the transition matrices for each partition: The AA substitution model must be the first entry in each line and must be separated by a comma from the gene name, just like the DNA token above. You can not assign different models of rate heterogeneity to different partitions, i.e. it will be either CAT, GAMMA, GAMMAI etc. for all partitions, as specified with -m. Finally, if you have a concatenated DNA and AA alignments, with DNA data at positions 1 - 500 and AA data at 501-1000 with the WAG model the partition file should look as follows: DNA, gene1 = 1-500 WAG gene2 = 501-1000 estimate_perpartbrlen Estimate individual per-partition branch lengths (-M) perl defined $partition perl ($value) ? " -M" : "" 0 The -M option switches on estimation of individual per-partition branch lengths. Only has effect when used in combination with -q and an alignment partition file. Branch lengths for individual partitions will be printed to separate files. A weighted average of the branch lengths is also computed by using the respective partition lengths (number of columns per partition). Note that, this does not take into account the "gappyness" of partitions, but I am currently not sure how to solve this problem. By default RAxML will compute a joined branch length estimate. specify_ML Specify an ML estimate of base frequencies (GTRGAMMA + X) X 2 invariable Estimate proportion of invariable sites (GTRGAMMA + I) I The invariable option is not recommended by the developer of RAxML. Please see the manual for details. perl $invariable 2 This option is not recommended by the developer of RAxML exclude_file Choose an input file that excludes the range of positions specifed in this file (-E) perl " -E excl" 2 excl This option is used to excludes specific positions in the matrix. For example, uploading a file that contains the text: 100-200 300-400 will create a file that excludes all columns between positions 100 and 200 as well as all columns between positions 300 and 400. Note that the boundary numbers (positions 100, 200, 300, and 400) will also be excluded. To exclude a single column write (100-100). This option does not run an analysis but just prints an alignment file without the excluded columns. Save this file to your data area, and then run the real analysis. If you use a mixed model, an appropriately adapted model file will also be written. The ntax element of the phylip files is automatically corrected Example: raxmlHPC -E excl -s infile -m GTRCAT -q part -n TEST. In this case the files with columns excluded will be named infile.excl and part.excl. set_weights Weight characters as specifed in this file (-a) perl " -a weights" 2 weights This option alows you to specify a column weight file name to assign individual weights to each column of the alignment. Those weights must be integers separated by any type and number of whitespaces within a separate file. There must, of course, be as many weights as there are columns in your alignment. The contents of an example weight file could look like this: 5 1 1 2 1 1 1 1 1 1 1 2 1 1 3 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 1 1 1 4 1 1 disable_seqcheck Disable checking for sequences with no values (-O) perl ($value) ? "-O" : "" 0 Please use the -O option with caution. It disables the check to see if sequences are completely devoid of information. See the RAxML manual for guidance perl $disable_seqcheck 54 mesquite_output Print output files that can be parsed by Mesquite. (-mesquite) perl $choose_bootstrap ne "x" && $choose_bootstrap ne "b" && !$altrun_number perl ($value) ? "--mesquite" : "" 0 54 nucleic_opts Nucleic Acid Options dna_gtrcat Choose model for bootstrapping phase perl $datatype eq "dna" || $datatype eq "rna" GTRCAT GTRGAMMA perl "-m $ascertainment$value$invariable$specify_ML" GTRCAT 2 Please choose a DNA model perl ($datatype eq "dna" || $datatype eq "rna") && $dna_gtrcat ne "GTRCAT" && $dna_gtrcat ne "GTRGAMMA" Sorry, the -f x option is valid only with GAMMA models perl $select_analysis eq "fx" && $dna_gtrcat ne "GTRGAMMA" The meaning of the model name GTRGAMMA used by RAxML 7.2.0 is exactly opposite that used in RAxML 7.0.4, so we have eliminated selection by model name. Instead we use a description of the model analysis. This selection gives GTR + Optimization of substitution rates + Optimization of site-specific evolutionary rates which are categorized into "numberOfCategories" distinct rate categories for greater computational efficiency. Final tree is evaluated under GTRGAMMA. GTRMIX and GTRCAT_GAMMA have been eliminated as options. FLOAT options that are native in RAxML 7.2.3 are currently not supported here. partitionUnder Evaluate DNA partitions only under this model perl $datatype eq "dna" HKY85 K80 JC69 perl (defined $value) ? " --$value":"" 2 This option specifies that all DNA partitions will evolve under the HKY85 model, this overrides all other model specifications for DNA partitions. Note that, the output of the program might look a bit weird, since unlike in the definition of the model, RAxML actually estimates the rates from A - G and C - T while all other rates are set to 1.0. Note that, this does not matter, since the rates in the rate matrix are relative rates; the results (likelihoods) will be the same protein_opts Protein Analysis Options prot_sub_model Choose GAMMA or CAT model: perl $datatype eq "protein" PROTGAMMA PROTCAT perl "-m $ascertainment$value$invariable$prot_matrix_spec$use_emp_freqs$use_ml_freqs$specify_ML" PROTCAT 2 Please choose a protein model perl $datatype eq "protein" && $prot_sub_model ne "PROTGAMMA" && $prot_sub_model ne "PROTCAT" Sorry, the -f x option is valid only with GAMMA models perl $select_analysis eq "fx" && $prot_sub_model ne "PROTGAMMA" prot_matrix_spec Protein Substitution Matrix perl $datatype eq "protein" DAYHOFF DCMUT JTT MTREV WAG RTREV CPREV VT BLOSUM62 MTMAM LG MTART MTZOA PMB HIVB HIVW JTTDCMUT FLU DUMMY DUMMY2 AUTO LG4M LG4X PROT_FILE GTR_UNLINKED GTR DAYHOFF Note: FLOAT and invariable sites (I) options are not exposed here. If you require this option, please contact mmiller@sdsc.edu. -m PROTCATmatrixName: analyses using the specified AA matrix + Optimization of substitution rates + Optimization of site-specific evolutionary rates which are categorized into numberOfCategories distinct rate categories for greater computational efficiency. Final tree might be evaluated automatically under PROTGAMMAmatrixName[f], depending on the tree search option. -m PROTGAMMAmatrixName[F] analyses use the specified AA matrix + Optimization of substitution rates + GAMMA model of rate heterogeneity (alpha parameter will be estimated) Available AA substitution models: DAYHOFF, DCMUT, JTT, MTREV, WAG, RTREV, CPREV, VT, BLOSUM62, MTMAM, LG, GTR. You can specify if you want to use empirical base frequencies. Please note that for mixed models you can in addition specify the per-gene AA model in the mixed model file (see manual for details). Also note that if you estimate AA GTR parameters on a partitioned dataset, they will be linked (estimated jointly) across all partitions to avoid over-parametrization. user_prot_matrix Upload a Custom Protein Substitution Matrix (-P) perl $datatype eq "protein" perl "-P Userproteinmatrix.txt" 2 Userproteinmatrix.txt Specify a file containing a user-defined Protein substitution model. This file must contain 420 entries, the first 400 entires are the AA substitution rates (this matrix must be symmetric) and the last 20 entries are the empirical base frequencies mulcustom_aa_matrices Use a Partition file that specifies AA Matrices perl $datatype eq "protein" Please choose a partition file specifying up to 5 partitions perl $mulcustom_aa_matrices && !defined $partition This option can be used to specify multiple custom matrices via a partition file. The filenames must be specified as firstpartition, secondpartition, thirdpartition, fourthpartition, and fifthpartition, in order, user_prot_matrixq1 Select the First Protein Substitution Matrix Called in Your Partition File perl $mulcustom_aa_matrices firstpartition This option allows the user to upload a Protein subsitution matrix user_prot_matrixq2 Select the Second Protein Substitution Matrix Called in Your Partition File perl $mulcustom_aa_matrices && defined $user_prot_matrixq1 secondpartition This option allows the user to upload a second Protein subsitution matrix user_prot_matrixq3 Select the Third Protein Substitution Matrix Called in Your Partition File perl $mulcustom_aa_matrices && defined $user_prot_matrixq2 thirdpartition This option allows the user to upload a third Protein subsitution matrix user_prot_matrixq4 Select the Fourth Protein Substitution Matrix Called in Your Partition File perl $mulcustom_aa_matrices && defined $user_prot_matrixq3 fourthpartition This option allows the user to upload a fourth Protein subsitution matrix user_prot_matrixq5 Select the Fifth Protein Substitution Matrix Called in Your Partition File perl $mulcustom_aa_matrices && defined $user_prot_matrixq4 fifthpartition This option allows the user to upload a fifth Protein subsitution matrix use_emp_freqs Use empirical frequencies? [F] perl $datatype eq "protein" F use_ml_freqs Make an ML estimate of frequencies [X] perl $datatype eq "protein" X Sorry you cant use bot the X and F optionsm, please choose one or the other perl $use_ml_freqs && $use_emp_freqs Sec_structure_opts RNA Secondary Structure Options sec_str_file perl $datatype eq "rna" Upload a Secondary Structure File (-S) sec_structure.txt perl (defined $value) ? " -S sec_structure.txt" : "" 2 This option allows you to provide a secondary structure file. The file can contain "." for alignment columns that do not form part of a stem and characters, while "(), [], and {}" are used to define stem regions and pseudoknots. rna_model Use an RNA Secondary Structure Substitution Model (-A) perl defined $sec_str_file S6A S6B S6C S6D S6E S7A S7B S7C S7D S7E S7F S16A S16B S16A perl "-A $value" 2 Use this option to specify one of the 6, 7, or 16 state RNA secondary structure substitution models.The nomenclature is identical to that used in the program PHASE. For more information, see PHASE documentation: 6 state model nomenclature: http://www.cs.manchester.ac.uk/ai/Software/phase/manual/node101.html; 7 state model nomenclature http://www.cs.manchester.ac.uk/ai/Software/phase/manual/node107.html; 16 state model nomenclature http://www.cs.manchester.ac.uk/ai/Software/phase/manual/node114.html bin_opts Binary Matrix Options bin_model Binary data model (-m) perl $datatype eq "binary" BINCAT BINGAMMA BINCAT perl "-m $ascertainment$value$invariable$specify_ML" 2 Please choose a binary model perl $datatype eq "binary" && $bin_model ne "BINCAT" && $bin_model ne "BINGAMMA" Sorry, the -f x option is valid only with GAMMA models perl $select_analysis eq "fx" && $bin_model ne "BINGAMMA" Binary data is handled in RAXML 7.2.0. Binary CAT use optimization of site-specific evolutionary rates, which are categorized into numberOfCategories (option -c) distinct rate categories for greater computational efficiency. Final tree might be evaluatedautomatically under BINGAMMA, depending on the tree search option. Binary GAMMA uses the GAMMA model of rate heterogeneity (alpha parameter will be estimated). The option for invariable sites is not provided at this time. The program's author supports the use of Gamma models. multi_opts Multiple State Morphological Matrix Options multi_model Multiple State Data Model (-m) perl $datatype eq "multi" MULTICAT MULTIGAMMA MULTICAT perl "-m $ascertainment$value$invariable$specify_ML" 2 Please choose a Multi-State model perl $datatype eq "multi" && $multi_model ne "MULTICAT" && $multi_model ne "MULTIGAMMA" Multi-state morphological data are handled in RAXML at V. 7.3.0 and above. Multi-state CAT uses optimization of site-specific evolutionary rates which are categorized into numberOfCategories distinct rate categories for greater computational efficiency. Final tree might be evaluated automatically under MULTIGAMMA depending on the tree search option Mutli-state GAMMA uses the GAMMA model of rate heterogeneity (alpha parameter will be estimated). Invariable sites (I) options are not exposed here. If you require this option, please contact mmiller@sdsc.edu. choose_multi_model Select a Multiple state data model (-K) perl $datatype eq "multi" ORDERED MK GTR GTR perl "-K $value" 2 Please choose a Multi-State data model perl $datatype eq "multi" && $choose_multi_model ne "ORDERED" && $choose_multi_model ne "MK" && $choose_multi_model ne "GTR" Multi-state morphological data are handled in RAXML 7.3.0 and above. Multi-state CAT uses optimization of site-specific evolutionary rates which are categorized into numberOfCategories distinct rate categories for greater computational efficiency. Final tree might be evaluated automatically under MULTIGAMMA depending on the tree search option Mutli-state GAMMA uses the GAMMA model of rate heterogeneity (alpha parameter will be estimated). The program's author supports the use of Gamma models. set_analysis Configure the Analysis select_analysis Select the Analysis Type fa fd fD fb fA fJ fe fg fG fh fT fx fk fE fu fv fo I J y fa "-f a" fd "" fD "-f D" fb "-f b" fA "-f A" fJ "-f J" fe "-f e" fg "-f g" fG "-f G" fh "-f h" fT "-f T" fx "-f x" fE "-f k" fE "fk" fu "-f u" fv "-f v" fo "-f o" I "-I $aposterior_bootstopping" J "-J $specify_mr" y "-y" fd To use the -f a option, please select Rapid Bootstrapping (-x) perl $choose_bootstrap ne "x" && $select_analysis eq "fa" To use the -f A option please specify a best tree with "-t" perl $select_analysis eq "fA" && !defined $treetop To use the -f J option, please specify a tree with "-t" perl $select_analysis eq "fJ" && !defined $treetop To use the -f e option, please specify a tree with "-t" perl $select_analysis eq "fe" && !defined $treetop To use the -f b option, please specify a best tree with "-t" and file containing multiple trees with the "-z" option perl $select_analysis eq "fb" && ( !defined $bunchotops || !defined $treetop) Sorry, you cannot compute a log likelihood test (-f g) with GTRCAT models, please use "GTRGAMMA for the bootstrapping phase and GTRGAMMA for the final tree" perl ($datatype eq "dna" || $datatype eq "rna" ) && $select_analysis eq "fg" && $dna_gtrcat eq "GTRCAT" To use the -f g option, please specify a file containing one or more trees with the "-z" option perl $select_analysis eq "fg" && !defined $bunchotops Sorry, you cannot compute a log likelihood test (-f G) with GTRCAT models, please use "GTRGAMMA for the bootstrapping phase and GTRGAMMA for the final tree" perl ($datatype eq "dna" || $datatype eq "rna" ) && $select_analysis eq "fG" && $dna_gtrcat eq "GTRCAT" To use the -f G option, please specify a file containing one or more trees with the "-z" option perl $select_analysis eq "fG" && !defined $bunchotops Sorry, you cannot compute a log likelihood test (-f h) with GTRCAT models, please use "GTRGAMMA for the bootstrapping phase and GTRGAMMA for the final tree" perl ($datatype eq "dna" || $datatype eq "rna" ) && $select_analysis eq "fh" && $dna_gtrcat eq "GTRCAT" To use the compute a log likelihood test option (-f h), please specify a best tree with "-t" and file containing multiple trees with the "-z" option perl $select_analysis eq "fh" && (!defined $bunchotops || !defined $treetop) Sorry, you cannot compute a log likelihood test (-f h) with GTRCAT models, please select PROTGAMMA for the model perl $datatype eq "protein" && $select_analysis eq "fh" && $prot_sub_model eq "PROTCAT" Sorry, you cannot compute a log likelihood test (-f h) with GTRCAT models, please select PROTGAMMA for the model perl $datatype eq "protein" && $select_analysis eq "fh" && $prot_sub_model eq "PROTCAT" Sorry, you cannot compute a log likelihood test (-f h) with GTRCAT models, please select BINGAMMA for the model perl $datatype eq "binary" && $select_analysis eq "fh" && $bin_model eq "BINCAT" Please specify a tree (via the -t option above) to use the -f T option perl $select_analysis eq "fT" && !defined $treetop Sorry, to use the -f k option you must provide a tree (-t) a partition (-q) and estimate per partition branch lengths (-M) option perl $select_analysis eq "fk" && (!defined $partition || !$estimate_perpartbrlen || !defined $treetop) Please specify a tree (via the -t option above) to use the -f u option perl $select_analysis eq "fu" && !defined $treetop Please specify a non-comprehensive reference tree (via the -t option above) to use the -f v option perl $select_analysis eq "fv" && !defined $treetop Sorry, you cannot use a posteriori bootstrapping with the -b or -x options perl $use_apobootstopping && ($choose_bootstrap eq "x" || $choose_bootstrap eq "b") In order to use the a posteriori bootstrapping option (-I), you must supply a file with topologies for a posteriori bootstopping (-z) perl $select_analysis eq "I" && !defined $aposterior_topologies && !defined $bunchotops Sorry, you cant use the -y option and provide a starting tree using -t. perl $select_analysis eq "y" && defined $treetop The -J option requires a tree file containing several UNROOTED trees as input. If you provide anything else, the job will not progress, but it will use all of the configured run time. Please be sure you use a correct input file. perl $select_analysis eq "J" -f d This is the default RAxML tree search algorithm and is substantially faster than the original search algorithm. It takes some shortcuts, but yields trees that are almost as good as the ones obtained from the full search algorithm. When -f b is specified, RAxML draws the bipartitions using a bunch of topologies (typically boot-strapped trees) specified with -z onto a single tree topology specified by -t (typically the best-scoring ML tree). When -f A is specified, RAxML computes marginal ancestral states/sequences on a given, fixed,and rooted reference tree. If you don't know what marginal ancestral states are please read Ziheng Yang's book on Computational Molecular Evolution.. The -f h option computes a log likelihood test (SHtest) between the best tree passed via -t and a bunch of other trees passed via -z. The model parameters will be estimated on the first tree only! The -f T option allows the user to do a more thorough tree search that uses the less lazy, i.e. more exhaustive SPR moves, in a stand alone mode. This algorithm is typically executed in the very end of a search done by -f a. If you use the -f o option you will typically get slightly better likelihood scores while the run times are expected to increase by factor 2 to 3. The -y option computes a randomized parsimony starting tree with RAxML and not execute an ML analysis of the tree specify -y. The program will exit after computation of the starting tree. This option can be useful if you want to assess the impact of randomized MP and Neighbor Joining starting trees on your search algorithm. They can also be used e.g. as starting trees for Derrick Zwickls GARLI program for ML inferences, which needs comparatively good starting trees to work well above approximately 500 taxa. specify_runs Specify the number alternative runs on distinct starting trees? (-N) perl $select_analysis ne "J" && $select_analysis ne "fg" && $select_analysis ne "fG" && $select_analysis ne "fh" && $select_analysis ne "fT" && $select_analysis ne "fE" && $select_analysis ne "fA" && $choose_bootstrap ne "x" && $choose_bootstrap ne "b" && $select_analysis ne "fb" && $select_analysis ne "y" && $select_analysis ne "fv" This option specifies the number of alternative runs on distinct starting trees. For example, if -N 10 is specfied, RAxML will compute 10 distinct ML trees starting from 10 distinct randomized maximum parsimony starting trees. altrun_number Enter number of number alternative runs (-N) perl $specify_runs perl "-N $value" 10 15 Please specify how many runs you wish to execute with the -N option (eg 10) perl $specify_runs && !defined $altrun_number Sorry, the value for alternative runs must 1000 or less for -N perl $altrun_number > 1000 if -N 10 is specfied, RAxML will compute 10 distinct ML trees starting from 10 distinct randomized maximum parsimony starting trees. bunchotops File with topologies for bipartitions (-z) perl ($select_analysis eq "fb" || $select_analysis eq "fh" || $select_analysis eq "fG" || $select_analysis eq "fg") && !defined $aposterior_topologies perl " -z topologies_file.tre" 2 topologies_file.tre The -z option is used in combination with the -f b,-f g, -f G, -f h,-f m,-f n options. The uploaded file should contain a number of trees in NEWICK format. The file should contain one tree per line without blank lines between trees. For example, you can directly read in a RAxML bootstrap result file. no_bfgs Don't use BFGS searching algorithm (--no-bfgs) 0 perl ($value)? "--no-bfgs":"" BFGS is a more efficient optimization algorithm for optimizing branch lengths and GTR parameters simultaneously. You can disable it using this option intermediate_treefiles Write intermediate tree files to a file (-j) 0 perl $select_analysis ne "g" && $select_analysis ne "G" && $select_analysis ne "fT" && $select_analysis ne "J" && $select_analysis ne "fA" && $select_analysis ne "fb" && $select_analysis ne "fE" && $select_analysis ne "y" && $select_analysis ne "fh" && $select_analysis ne "fx" && $select_analysis ne "fu" && $select_analysis ne "fv" 2 perl ($value)?" -j ":"" This will simply print out a couple of intermediate trees during the tree search and not the final tree only. The intermediate trees are written to files called: RAxML_checkpoint.TEST.0, RAxML_checkpoint.TEST.1, etc. convergence_criterion Use ML search convergence criterion. (-D) 0 perl $select_analysis ne "fg" && $select_analysis ne "fG" && $select_analysis ne "fE" && $select_analysis ne "J" && $select_analysis ne "fA" && $select_analysis ne "fb" && $select_analysis ne "y" && $select_analysis ne "fh" && $select_analysis ne "fx" && !$specify_runs && $select_analysis ne "fu" && $select_analysis ne "fv" 2 perl ($value)?" -D ":"" The tree search convergence criterion "-D" has no effect in conjunction with the: "-x" or "-f a" options. perl $value && ($select_analysis eq "fa" || $choose_bootstrap eq "x" ) -D option = ML search convergence criterion. This will break off ML searches if the relative RobinsonFoulds distance between the trees obtained from two consecutive lazy SPR cycles is smaller or equal to 1%. Usage recommended for very large datasets in terms of taxa. On trees with more than 500 taxa this will yield execution time improvements of approximately 50% while yielding only slightly worse trees. specify_mr Specify majority rule consensus tree (-J) technique perl $select_analysis eq "J" MR MRE STRICT MR_DROP STRICT_DROP MR Please select a majority rule option for the -J option perl !$specify_mr You must use a collection of trees as your input file for this option. The option lets you compute a majority rule consensus tree with "MR" or extended majority rule consensus tree with "J MRE" or strict consensus tree with "J STRICT". Options "J STRICT_DROP" and "J MR_DROP" will execute an algorithm that identifies dropsets which contain rogue taxa as proposed by Pattengale et al. in the paper "Uncovering hidden phylogenetic consensus". ascertainment_config Ascertainment Bias Confguration ascertainment Correct for Ascertainment bias (ASC_) perl !$invariable ASC_ 2 Ascertainment bias correction will be applied only to partitions for which it is requested in the partition file (-q) perl $ascertainment eq "ASC_" && defined $partition This is useful for binary/morphological datasets that only contain variable sites (the identical morphological features are usually not included in the alignments, hence you need to correct for this, see, e.g., http://sysbio.oxfordjournals.org/content/50/6/913.short).For DNA data this option might be useful when you analyze alignments of SNPs that also don't contain constant sites. Note that, for mathematical and numerical reasons you can not apply an ascertainment bias correction to datasets or partitions that contain constantsites. In this case, RAxML will exit with an error. ascertainment_corr Ascertainment bias correction type (--asc-corr) perl "--asc-corr $value" lewis felsenstein stamatakis 40 To use the Felsentein option (--asc-corr), you must specify the number of invariable sites in a file using -q perl $ascertainment_corr eq "felsenstein" && !defined $partition To use the Stamatakis option (--asc-corr), you must specify the number of invariable sites per state for each partition in a file using -q perl $ascertainment_corr eq "stamatakis" && !defined $partition Please specify the ascertainment correction method perl $ascertainment eq "ASC_" && !defined $ascertainment_corr This option allows to specify the type of ascertainment bias correction you wish to use. There are three types available: Lewis: the standard correction by Paul Lewis, Felsenstein: a correction introduced by Joe Felsenstein that allows to explicitely specify the number of invariable sites (if known) one wants to correct for. Stamatakis: a correction introduced by myself that allows to explicitly specify the number of invariable sites for each character (if known) one wants to correct for. Flesenstein and Stamatkis corrections are accompanied by an upload file specified by the -q option, even if only one partiion is present. For file formatting, please see the RaxML 8.1 or higher manual. ascertainment_pfile1 Choose Ascertainment bias correction file 1 (will be named p1.txt) perl $ascertainment_corr eq "felsenstein" || $ascertainment_corr eq "stamatakis" p1.txt 40 To use the Felsentein option (--asc-corr), you must specify the number of invariable sites in a file perl $ascertainment_corr eq "felsenstein" && !defined $ascertainment_pfile1 To use the Stamatakis option (--asc-corr), you must provide the number of invariable sites per state in a file perl $ascertainment_corr eq "stamatakis" && !defined $ascertainment_pfile1 Ascertainment corrections that follow the Stamatakis or Felsenstein models require files that specify the number of invariant sites per state or the number of invariable sites, respectively. This file will be named p1.txt, For information on file formatting, please see the RaxML 8.1 or higher manual. ascertainment_pfile2 Choose Ascertainment bias correction file 2 (will be named p2.txt) perl defined $ascertainment_pfile1 p2.txt 40 This file will be named p2.txt, For information on file formatting, please see the RaxML 8.1 or higher manual. ascertainment_pfile3 Choose Ascertainment bias correction file 3 (will be named p3.txt) perl defined $ascertainment_pfile2 p3.txt 40 This file will be named p3.txt, For information on file formatting, please see the RaxML 8.1 or higher manual. ascertainment_pfile4 Choose Ascertainment bias correction file 4 (will be named p4.txt) perl defined $ascertainment_pfile3 p4.txt 40 This file will be named p4.txt, For information on file formatting, please see the RaxML 8.1 or higher manual. ascertainment_pfile5 Choose Ascertainment bias correction file 5 (will be named p5.txt) perl defined $ascertainment_pfile4 p5.txt 40 This file will be named p5.txt, For information on file formatting, please see the RaxML 8.1 or higher manual. ascertainment_pfile6 Choose Ascertainment bias correction file 6 (will be named p6.txt) perl defined $ascertainment_pfile5 p6.txt 40 This file will be named p6.txt, For information on file formatting, please see the RaxML 8.1 or higher manual. bootstrap_config Configure Bootstrapping choose_bootstrap Choose a Bootstrapping Type perl $select_analysis eq "fd" || $select_analysis eq "fa" || $select_analysis eq "fo" b x The -b option allows you to turn on non-parametric bootstrapping. Note that parallel bootstraps with the parallel version raxmlHPC-MPI are not reproducible despite the fact that you specify a random number seed. Use rapid bootstrapping (-x) to turn on rapid bootstrapping. CAUTION: unlike in previous versions of RAxML will conduct rapid BS replicates under the model of rate heterogeneity you specified via m and not by default under CAT seed_value Enter a random seed value for bootstrapping 12345 perl $choose_bootstrap eq "b" || $choose_bootstrap eq "x" 2 Please enter a random seed for the -b option (eg 12345) perl $choose_bootstrap eq "b" && !defined $seed_value Please enter a random seed for the -x option (eg 12345) perl !defined $seed_value && $choose_bootstrap eq "x" This random number is provided to assure that there is comparability between runs. mulparambootstrap_seed_val perl $choose_bootstrap eq "b" perl " -b $seed_value" 12345 2 bootstrap_seed_val Enter a random seed value for bootstrapping perl $choose_bootstrap eq "x" perl " -x $seed_value" 12345 2 printbrlength Print branch lengths (-k) perl ($value)?" -k":"" 0 2 The -k option causes bootstrapped trees to be printed with branch lengths. The bootstraps will require a bit longer to run under this option because model parameters will be optimized at the end of each run under GAMMA or GAMMA+P-Invar respectively. choose_bootstop Specify bootstrap protocol perl $choose_bootstrap eq "x" || $choose_bootstrap eq "b" specify bootstop specify Please select "Specify an explicit number of bootstraps" or "Let RaxML halt bootstrapping automatically" perl !defined $choose_bootstop This option instructs Raxml to automatically halt bootstrapping when certain criteria are met, instead of specifying the number of bootstraps for an analysis. The exact criteria are specified/configured using subsequent entry fields. bootstrap_value Bootstrap iterations (-N) perl $choose_bootstop eq "specify" perl " -N $value" 100 2 Please enter number of bootstraps desired (-N) (eg 100) perl $choose_bootstop eq "specify" && !defined $bootstrap_value Sorry, the number of bootstraps cannot exceed 1,000 (-N) perl $bootstrap_value > 1000 Specifies the number of alternative runs on distinct starting trees. If 10, RAxML computes 10 distinct ML trees starting from 10 distinct randomized maximum parsimony starting trees. In combination with the Random seed for rapid bootstrap (-x) invoke a rapid BS analysis. bootstopping_type Select Bootstopping Criterion: (autoMRE is recommended) perl $choose_bootstop eq "bootstop" perl "-N $value" autoFC autoMR autoMRE autoMRE_IGN autoMRE Please choose a bootstopping criterion perl !defined $bootstopping_type aposterior_bootstopping Select the criterion for a posteriori bootstopping analysis (-I) perl $select_analysis eq "I" autoFC autoMR autoMRE autoMRE_IGN autoMRE This option allows the user to conduct a posteriori bootstopping analysis based on a set of bootstrapped trees. Use: autoFC for the frequency-based criterion, autoMR for the majority-rule consensus tree criterion, autoMRE for the extended majority-rule consensus tree criterion and autoMRE_IGN for metrics similar to MRE, but include bipartitions under the threshold whether they are compatible or not. This emulates MRE but is faster to compute. For any of these options, you also need to upload a tree file containing several bootstrap replicates via "-z" aposterior_topologies File with topologies for a posteriori bootstopping (-z) perl $select_analysis eq "I" && !defined $bunchotops perl " -z apotopologies_file.tre" 2 apotopologies_file.tre all_outputfiles *