RAxML-HPC2 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 raxmlhpc2_tgb raxmlhpc_hybridlogic2 perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $specify_bootstraps && $bootstop < 50 && !$more_memory 0 raxmlhpc_hybridlogic2_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=6\\n" . "threads_per_process=4\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $specify_bootstraps && $bootstop < 50 && !$more_memory 0 raxmlhpc_hybridlogic2b perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $specify_runs && $altrun_number < 50 && !$more_memory 0 raxmlhpc_hybridlogic2b_scheduler scheduler.conf perl $specify_runs && $altrun_number < 50 && !$more_memory perl "jobtype=mpi\\n" . "mpi_processes=6\\n" . "threads_per_process=4\\n" . "node_exclusive=1\\n" . "nodes=1\\n" 0 raxmlhpc_hybridlogic3 perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $specify_bootstraps && $runtime > 2 && $bootstop >= 50 && !$more_memory 0 raxmlhpc_hybridlogic3_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=12\\n" . "threads_per_process=4\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $specify_bootstraps && $runtime > 2 && $bootstop >= 50 && !$more_memory 0 raxmlhpc_hybridlogic3s perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $specify_bootstraps && $runtime <= 2 && $bootstop >= 50 && !$more_memory 0 raxmlhpc_hybridlogic3s_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=6\\n" . "threads_per_process=4\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $specify_bootstraps && $runtime <= 2 && $bootstop >= 50 && !$more_memory 0 raxmlhpc_hybridlogic4 perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $specify_runs && $runtime > 2 && $altrun_number >= 50 && !$more_memory 0 raxmlhpc_hybridlogic4_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=12\\n" . "threads_per_process=4\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $specify_runs && $runtime > 2 && $altrun_number >= 50 && !$more_memory 0 raxmlhpc_hybridlogic4s perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $specify_runs && $runtime <= 2 && $altrun_number >= 50 && !$more_memory 0 raxmlhpc_hybridlogic4s_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=6\\n" . "threads_per_process=4\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $specify_runs && $runtime <= 2 && $altrun_number >= 50 && !$more_memory 0 raxmlhpc_hybridlogic11 perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $use_bootstopping && $runtime > 2 && !$more_memory 0 raxmlhpc_hybridlogic11_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=12\\n" . "threads_per_process=4\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $use_bootstopping && $runtime > 2 && !$more_memory 0 raxmlhpc_hybridlogic11s perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $use_bootstopping && $runtime <= 2 && !$more_memory 0 raxmlhpc_hybridlogic11s_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=6\\n" . "threads_per_process=4\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $use_bootstopping && $runtime <= 2 && !$more_memory 0 raxmlhpc_hybridlogic13 perl "raxmlHPC-PTHREADS_8.2.12_centos7 " perl !$mulparambootstrap_seed && !$bootstrap_seed && !$specify_runs && !$more_memory 0 raxmlhpc_hybridlogic13b_scheduler scheduler.conf perl "threads_per_process=12\\n" . "node_exclusive=0\\n" . "nodes=1\\n" perl !$mulparambootstrap_seed && !$bootstrap_seed && !$specify_runs && !$more_memory 0 raxmlhpc_hybridlogic22 perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $runtime > 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 20*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 30*(1024*1024*1024) 0 raxmlhpc_hybridlogic22_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=8\\n" . "threads_per_process=6\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $runtime > 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 20*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 30*(1024*1024*1024) 0 raxmlhpc_hybridlogic23 perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime > 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 30*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 40*(1024*1024*1024) 0 raxmlhpc_hybridlogic23_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=6\\n" . "threads_per_process=8\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $runtime > 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 30*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 40*(1024*1024*1024) 0 raxmlhpc_hybridlogic24 perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime > 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 40*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 60*(1024*1024*1024) 0 raxmlhpc_hybridlogic24_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=4\\n" . "threads_per_process=12\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $runtime > 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 40*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 60*(1024*1024*1024) 0 raxmlhpc_hybridlogic25 perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime > 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 60*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 120*(1024*1024*1024) 0 raxmlhpc_hybridlogic25_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=2\\n" . "threads_per_process=24\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $runtime > 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 60*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 120*(1024*1024*1024) 0 raxmlhpc_hybridlogic22s perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $runtime <= 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 20*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 30*(1024*1024*1024) 0 raxmlhpc_hybridlogic22_scheduler_s scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=4\\n" . "threads_per_process=6\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $runtime <= 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 20*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 30*(1024*1024*1024) 0 raxmlhpc_hybridlogic23s perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime <= 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 30*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 40*(1024*1024*1024) 0 raxmlhpc_hybridlogic23_scheduler_s scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=3\\n" . "threads_per_process=8\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $runtime <= 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 30*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 40*(1024*1024*1024) 0 raxmlhpc_hybridlogic24s perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime <= 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 40*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 60*(1024*1024*1024) 0 raxmlhpc_hybridlogic24_scheduler_s scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=2\\n" . "threads_per_process=12\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $runtime <= 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 40*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 60*(1024*1024*1024) 0 raxmlhpc_hybridlogic25s perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime <= 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 60*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 120*(1024*1024*1024) 0 raxmlhpc_hybridlogic25_scheduler_s scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=1\\n" . "threads_per_process=24\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $runtime <= 2 && $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) > 60*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*16) < 120*(1024*1024*1024) 0 raxmlhpc_hybridlogic27 perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $runtime > 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 20*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 30*(1024*1024*1024) 0 raxmlhpc_hybridlogic27_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=8\\n" . "threads_per_process=6\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $runtime > 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 20*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 30*(1024*1024*1024) 0 raxmlhpc_hybridlogic28 perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime > 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 30*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 40*(1024*1024*1024) 0 raxmlhpc_hybridlogic28_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=6\\n" . "threads_per_process=8\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $runtime > 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 30*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 40*(1024*1024*1024) 0 raxmlhpc_hybridlogic29 perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime > 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 40*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 60*(1024*1024*1024) 0 raxmlhpc_hybridlogic29_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=4\\n" . "threads_per_process=12\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $runtime > 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 40*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 60*(1024*1024*1024) 0 raxmlhpc_hybridlogic30 perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime > 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 60*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 120*(1024*1024*1024) 0 raxmlhpc_hybridlogic30_scheduler scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=2\\n" . "threads_per_process=24\\n" . "node_exclusive=1\\n" . "nodes=2\\n" perl $runtime > 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 60*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 120*(1024*1024*1024) 0 raxmlhpc_hybridlogic27s perl "raxmlHPC-HYBRID_8.2.12_centos7 " perl $runtime <= 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 20*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 30*(1024*1024*1024) 0 raxmlhpc_hybridlogic27_scheduler_s scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=4\\n" . "threads_per_process=6\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $runtime <= 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 20*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 30*(1024*1024*1024) 0 raxmlhpc_hybridlogic28s perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime <= 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 30*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 40*(1024*1024*1024) 0 raxmlhpc_hybridlogic28_scheduler_s scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=3\\n" . "threads_per_process=8\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $runtime <= 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 30*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 40*(1024*1024*1024) 0 raxmlhpc_hybridlogic29s perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime <= 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 40*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 60*(1024*1024*1024) 0 raxmlhpc_hybridlogic29_scheduler_s scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=2\\n" . "threads_per_process=12\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $runtime <= 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 40*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 60*(1024*1024*1024) 0 raxmlhpc_hybridlogic30s perl "raxmlHPC-HYBRID_8.2.12_centos7" perl $runtime <= 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 60*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 120*(1024*1024*1024) 0 raxmlhpc_hybridlogic30_scheduler_s scheduler.conf perl "jobtype=mpi\\n" . "mpi_processes=1\\n" . "threads_per_process=24\\n" . "node_exclusive=1\\n" . "nodes=1\\n" perl $runtime <= 2 && $more_memory && $datatype eq "protein" && (($ntax - 2)*($nchar)*8*80) > 60*(1024*1024*1024) && (($ntax - 2)*($nchar)*8*80) < 120*(1024*1024*1024) 0 infile Sequences File (relaxed phylip format) (-s) 1 infile.txt infile_regular Sequences File (relaxed phylip format) (-z) perl !$compute_mr perl "-s infile.txt" 1 infile_Joption Tree file (-z) perl $compute_mr 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 168 or less perl $runtime > 168 Maximum Hours to Run must be greater than 0.1 perl $runtime < 0.1 perl "runhours=$value\\n" 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 $specify_bootstraps && $bootstop < 50 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 $specify_runs && $altrun_number < 50 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 $specify_runs && $runtime <= 2 && $altrun_number >= 50 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 $specify_runs && $runtime > 2 && $altrun_number >= 50 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 $specify_bootstraps && $runtime <= 2 && $bootstop >= 50 && !$ancestral_states && !$fast_tree && !$bipartitions && !$startingtreeonly && !$log_likelihood && !$morpho_weight_calibration && !$classify_into_reftree && !$compute_mr 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 $specify_bootstraps && $runtime > 2 && $bootstop >= 50 && !$ancestral_states && !$fast_tree && !$bipartitions && !$startingtreeonly && !$log_likelihood && !$morpho_weight_calibration && !$classify_into_reftree && !$compute_mr 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 $use_bootstopping && $runtime <= 2 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 $use_bootstopping && $runtime > 2 The job will run on 12 processors as configured. If it runs for the entire configured time, it will consume 12 x $runtime cpu hours perl !$mulparambootstrap_seed && !$bootstrap_seed && !$specify_runs && !$more_memory 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. outsuffix Set a name for output files perl "-n $value" result Please enter a name for the output files perl !defined $outsuffix 1 MLsearch_CAT Enable ML searches under CAT (-F) perl !$bootstrap_seed && !$mulparambootstrap_seed && !$mlsearch perl ($value)? " -F " : "" 0 This option allows one to enable tree searches under CAT for veyr 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. more_memory I have a data set that may require more than 20 GB of memory perl $datatype eq "dna" || $datatype eq "protein" 0 To figure out if you need this option, use the RAxML memory calculator here: http://sco.h-its.org/exelixis/software.html. Input the number of taxa, and patterns (or characters), and the gamma model(AA or DNA, as appropriate for your data). Multiply the value you receive times 32. If that value is greater than 64 GB, then you should select this option. If you dont, the run will fail silently at the end due to an out of memory error. nchar Enter the number of patterns in your dataset perl $more_memory Please enter a value for the number of patterns in your data matrix perl !defined $nchar The number of patterns in the matrix must 1 or greater. perl $nchar < 1 15 Knowing the number of characters in your dataset helps us determine the most efficient way to run raxml. We need to know the number of characters per row in the input data matrix. ntax Enter the number of taxa in your dataset perl $more_memory Please enter a value for the number of taxa in your data matrix perl !defined $ntax The number of taxa in the matrix must 1 or greater. perl $nchar < 1 Your job does not require the large memory option, please uncheck the large memory box and run normally. perl $more_memory && $datatype eq "dna" &&(($ntax - 2)*($nchar)*8*16) < 20*(1024*1024*1024) Your job will probably not complete as configured, please contact us for help. perl $more_memory && $datatype eq "dna" && (($ntax-2)* ($nchar*8*16)) > 120*(1024*1024*1024) Your job does not require the large memory option, please uncheck the large memory box and run normally. perl $more_memory && $datatype eq "protein" && (($ntax-2) * ($nchar*8*80)) < 20*(1024*1024*1024) Your job will probably not complete as configured, please contact us for help. perl $more_memory && $datatype eq "protein" && (($ntax-2)* ($nchar*8*80)) > 120*(1024*1024*1024) 15 Knowing the number of taxa and patterns in your dataset helps us determine the most efficient way to run raxml. datatype Please select the Data Type protein dna rna binary multi dna 2 outgroup Outgroup (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" && $bin_model eq "BINCAT") This option allows you to disable rate heterogeneity anong the sites. Valid for CAT model only. treetop Supply a tree (Not available when doing rapid bootstrapping, -x) (-t) perl !$bootstrap_seed 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) 1 Please provide a parsimony seed perl $specify_runs && !defined $parsimony_seed_val Please provide a parsimony seed perl $startingtreeonly && !defined $parsimony_seed_val Specify a random number seed. The -p option allows you and others to reproduce your results (reproducible/verifiable experiments) and will help Alexis debug the program. Do not use this option if you want to generate multiple different starting trees. parsimony_seed_val Enter a random seed value for parsimony inferences (gives reproducible results from random starting tree) perl ($value) ? " -p $value" : "" 12345 2 perl $provide_parsimony_seed 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 (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 $binary_backbone && !$startingtreeonly 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. ascertainment Correct for Ascertainment bias (ASC_) perl !$invariable ASC_ 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 2 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 constant sites. In this case, RAxML will exit with an error. Ascertainment bias correction will be applied only to partitions for which it is requested in the partition file (-q). For example, even if the command line states -m ASC_GTRGAMMA, if the partition file does not specify ASC_ for the DNA partition, the ASC_ prefix will be ignored. ascertainment_corr Ascertainment bias correction type (--asc-corr) perl "--asc-corr $value" lewis felsenstein stamatakis lewis 40 To use the Felsentein option, 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, 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 select an ascertainment correction method perl !defined $ascertainment_corr && $ascertainment 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. 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 with caution. It disables the check to see if sequences are completely devoid of information. See the RAxML manual for guidance perl $disable_seqcheck 54 mesquite_output Print output files that can be parsed by Mesquite. (-mesquite) perl ($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$invariable" GTRCAT 2 Please choose a DNA model perl ($datatype eq "dna" || $datatype eq "rna") && $dna_gtrcat ne "GTRCAT" && $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. 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" 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 $compute_ml_distances && $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 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? perl $datatype eq "protein" F 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" 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 $compute_ml_distances && $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" 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" && $choose_multi_model ne "GTR" Multi-state morphological data are handled in RAXML 7.3.0 and above. Multi-state CAT uses optimization of site-specific evolutionary rates which are categorized into numberOfCategories distinct rate categories for greater computational efficiency. Final tree might be evaluated automatically under MULTIGAMMA depending on the tree search option Mutli-state GAMMA uses the GAMMA model of rate heterogeneity (alpha parameter will be estimated). The program's author supports the use of Gamma models. set_analysis Select the Analysis startingtreeonly Only compute a randomized parsimony starting tree (-y) perl ($value)?" -y":"" 0 perl !$compute_mr && !$thorough_opt && !$ancestral_states && !$fast_tree && !$mulparambootstrap_seed &&!$bootstrap_seed && !$bipartitions && ! defined $constraint && !$log_likelihood && !$compute_ml_distances && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree 2 If you want to only compute 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 !$compute_mr && !$thorough_opt && !$fast_tree && !$ancestral_states && !$bootstrap_seed && !$mulparambootstrap_seed && !$bipartitions && !$startingtreeonly && !$log_likelihood && !$classify_into_reftree 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 perl $specify_runs perl "-N $value" 15 Please specify how many runs you wish to execute (eg 10) perl $specify_runs && !defined $altrun_number Sorry, the value for alternative runs must 1000 or less perl $altrun_number > 1000 if -N 10 is specfied, RAxML will compute 10 distinct ML trees starting from 10 distinct randomized maximum parsimony starting trees. no_bfgs Don't use BFGS searching algorithm (--no-bfgs) 0 perl ($value)? "--no-bfgs":"" BFGS is a more efficient optimization algorithm for optimizing branch lengths and GTR parameters simultaneously. YUOu can disable it using this option bipartitions Draw bipartitions onto a single tree topology. (-f b) perl ($value)?" -f b ":"" 0 2 perl !$compute_mr && !$thorough_opt && !$ancestral_states && !$fast_tree && !$mulparambootstrap_seed &&!$bootstrap_seed && !$startingtreeonly && !$log_likelihood && !$compute_ml_distances && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree To use the -f b option you must specify a best tree with "-t" and file containing multiple trees with the "-z" option perl !$bootstrap_seed && $bipartitions && ( !defined $bunchotops || !defined $treetop) When this is specified, RAxML draws the bipartitions using a bunch of topologies (typically boot-strapped trees) specified with -z onto a single tree topology specified by -t (typically the best-scoring ML tree). ancestral_states Compute Marginal Ancestral States using a rooted reference tree. (-f A) perl ($value)?" -f A ":"" 0 2 perl !$compute_mr && !$thorough_opt && !$bipartitions && !$fast_tree && !$mulparambootstrap_seed &&!$bootstrap_seed && !$startingtreeonly && !$log_likelihood && !$compute_ml_distances && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree To use the -f A option you must specify a best tree with "-t" perl !$bootstrap_seed && $ancestral_states && !defined $treetop When this is specified, RAxML draws the bipartitions using a bunch of topologies (typically boot-strapped trees) specified with -z onto a single tree topology specified by -t (typically the best-scoring ML tree). log_likelihood Compute a log likelihood test (-f h) perl ($value)?" -f h ":"" 0 2 perl !$compute_mr && !$thorough_opt && !$ancestral_states && !$fast_tree && !$mulparambootstrap_seed && !$bootstrap_seed && !$startingtreeonly && !$bipartitions && !$compute_ml_distances && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree To use the "compute a log likelihood test option" you must specify a file containing multiple trees with the "-z" option perl $log_likelihood && !defined $bunchotops To use the "compute a log likelihood test option" you must specify a best tree with "-t" perl $log_likelihood && !defined $bunchotops Sorry, you cannot compute a log likelihood test with GTRCAT models, please use "GTRGAMMA for the bootstrapping phase and GTRGAMMA for the final tree" perl ($datatype eq "dna" || $datatype eq "rna" ) && $log_likelihood && $dna_gtrcat eq "GTRCAT" Sorry, you cannot compute a log likelihood test with GTRCAT models, please select PROTGAMMA for the model perl $datatype eq "protein" && $log_likelihood && $prot_sub_model eq "PROTCAT" Sorry, you cannot compute a log likelihood test with GTRCAT models, please select PROTGAMMA for the model perl $datatype eq "protein" && $log_likelihood && $prot_sub_model eq "PROTCAT" Sorry, you cannot compute a log likelihood test with GTRCAT models, please select BINGAMMA for the model perl $datatype eq "binary" && $log_likelihood && $bin_model eq "BINCAT" When this is specified, RAxML will compute a log likelihood test (SH-test [21]) between a best tree passed via -t and a bunch of other trees passed via -z. Example: raxmlHPC -f h -t ref -z trees -s alg -m GTRGAMMA -n TEST thorough_opt Do A Final Opimization of ML Tree (-f T) 0 perl !$compute_mr && !$ancestral_states && !$bipartitions && !$fast_tree && !$mulparambootstrap_seed &&!$bootstrap_seed && !$startingtreeonly && !$log_likelihood && !$compute_ml_distances && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree 2 perl ($value)?" -f T ":"" You must specify a tree (via the -t option above) to use the -f T option perl $thorough_opt && !defined $treetop Sorry, you cant use the -f T option with the high memory option perl $thorough_opt && $more_memory The -f T option allows the user to do a more thorough tree search that uses the less lazy, i.e. more exhaustive SPR moves, in a stand alone mode. This algorithm is typically executed in the very end of a search done by -f a. intermediate_treefiles Write intermediate tree files to a file (-j) 0 perl !$compute_mr && !$ancestral_states && !$bipartitions && !$fast_tree && !$startingtreeonly && !$log_likelihood && !$compute_ml_distances && !$morpho_weight_calibration && !$classify_into_reftree 2 perl ($value)?" -j ":"" This will simply print out a couple of intermediate trees during the tree search and not the final tree only. The intermediate trees are written to files called: RAxML_checkpoint.TEST.0, RAxML_checkpoint.TEST.1, etc. convergence_criterion Use ML search convergence criterion. (-D) 0 perl !$compute_mr && !$ancestral_states && !$bipartitions && !$startingtreeonly && !$log_likelihood && !$compute_ml_distances && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree 2 perl ($value)?" -D ":"" The tree search convergence criterion "-D" has no effect in conjunction with the: "-x" or "-f a" options. perl $value && $bootstrap_seed -D option = ML search convergence criterion. This will break off ML searches if the relative RobinsonFoulds distance between the trees obtained from two consecutive lazy SPR cycles is smaller or equal to 1%. Usage recommended for very large datasets in terms of taxa. On trees with more than 500 taxa this will yield execution time improvements of approximately 50% while yielding only slightly worse trees. compute_mr Compute majority rule consensus tree (-J) perl ($dna_gtrcat eq "GTRGAMMA" || $prot_sub_model eq "PROTGAMMA" || $bin_model eq "BINGAMMA" || $multi_model eq "MULTIGAMMA" ) && !$compute_ml_distances && !$thorough_opt && !$ancestral_states && !$fast_tree && !$mulparambootstrap_seed && !$bootstrap_seed && !$bipartitions && !$startingtreeonly && !$log_likelihood && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree 0 This option allows the user to compute majority rule consensus tree or extended majority rule consensus tree from an uploaded file containing several trees (-z) specify_mr Specify majority rule consensus tree (-J) technique perl $compute_mr perl "-J $value" MR MRE STRICT MR_DROP STRICT_DROP MR Please select a majority rule option perl !$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". bunchotops File with topologies for bipartitions or bootstopping (-z) perl " -z topologies_file.tre" 2 topologies_file.tre perl ($bipartitions || $log_likelihood) && !defined $apo_tops The -z option is used in combination with the -f b,-f h,-f m,-f n options. The uploaded file should contain a number of trees in NEWICK format. The file should contain one tree per line without blank lines between trees. For example, you can directly read in a RAxML bootstrap result file. compute_ml_distances Compute pair-wise ML distances (-f x; GAMMA models only) 0 perl !$compute_mr && !$thorough_opt && !$ancestral_states && !$fast_tree && !$mulparambootstrap_seed && !$bootstrap_seed && !$bipartitions && !$startingtreeonly && !$log_likelihood && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree perl ($value)?" -f x ":"" Sorry, the -f x option is valid only with GAMMA models perl $compute_ml_distances && $dna_gtrcat ne "GTRGAMMA" Compute pair-wise ML distances, ML model parameters will be estimated on an MP starting tree or a user-defined tree passed via "-t". fast_tree Run very fast experimental tree search(-f E) 0 perl !$compute_mr && !$thorough_opt && !$ancestral_states && !$compute_ml_distances && !$mulparambootstrap_seed && !$bootstrap_seed && !$bipartitions && !$startingtreeonly && !$log_likelihood && !$specify_runs && !$morpho_weight_calibration && !$classify_into_reftree perl ($value)?" -f E ":"" Any constraint trees will be ignored perl $fast_tree Compute pair-wise ML distances, ML model parameters will be estimated on an MP starting tree or a user-defined tree passed via "-t". morpho_weight_calibration Execute morphological weight calibration using maximum likelihood (-f u) 0 perl ($datatype eq "binary" || $datatype eq "multi")&&(!$classify_into_reftree && !$thorough_opt && !$ancestral_states && !$fast_tree && !$mulparambootstrap_seed && !$bootstrap_seed && !$bipartitions && !$startingtreeonly && !$log_likelihood) && !$compute_mr perl ($value)?" -f u ":"" You must specify a starting tree (via the -t option above) to use the -f x option perl $morpho_weight_calibration && !defined $treetop This option will determine to which degree the sites of a morphological alignment are congruent with a given molecular reference tree. As output it will generate a RAxML weight file that reflects the degree of congruence and that can be subsequently read via the a option to conduct a more informed evolutionary placement of fossils using the evolutionary placement algorithm (f -v option). For more details, see the corresponding papers: Evolutionary placement http://sysbio.oxfordjournals.org/content/60/3/291.short Fossil calibration http://sco.h-its.org/exelixis/pubs/Exelixis-RRDR-2009-1.pdf classify_into_reftree Classify a bunch of environmental sequences into a reference tree using thorough read insertions(-f v) 0 perl !$compute_mr && !$morpho_weight_calibration && !$thorough_opt && !$ancestral_states && !$fast_tree && !$mulparambootstrap_seed && !$bootstrap_seed && !$bipartitions && !$startingtreeonly && !$log_likelihood && !$specify_runs perl ($value)?" -f v ":"" You must specify a starting tree (via the -t option above) to use the -f x option perl $classify_into_reftree && !defined $treetop Classify a bunch of environmental sequences into a reference tree using thorough read insertions you will need to start RAxML with a noncomprehensive reference tree and an alignment containing all sequences (reference + query) For details on the design and performance of this algorithm, see: http://sysbio.oxfordjournals.org/content/60/3/291.short. RAxML will produce a couple of idiosyncratic output files for the placements, but also an output file according to the common file standard defined by Erick Matsen for his pplacer(see http://matsen.fhcrc.org/pplacer/) program that is similar to the EPA (Evolutionary Placement Algorithm) and myself. The common file format is described in this paper here: http://www.plosone.org/article/info %3Adoi%2F10.1371%2Fjournal.pone.0031009 bootstrap_config Configure Bootstrapping mulparambootstrap_seed Conduct Multiparametric Bootstrapping? (-b) 0 perl !$compute_mr && !$thorough_opt && !$ancestral_states && !$fast_tree && !$bootstrap_seed && !$startingtreeonly && !$compute_ml_distances && !$bipartitions && !$log_likelihood && !$compute_mr && !$specify_runs This option allows you to turn on non-parametric bootstrapping. To allow for reproducibility of runs in the sequential program, you have to specify a random number seed. mulparambootstrap_seed_val Enter a random seed value for multi-parametric bootstrapping perl ($value) ? " -b $value" : "" 12345 perl $mulparambootstrap_seed 2 Please enter a random seed for the -b option (eg 12345) perl $mulparambootstrap_seed && !defined $mulparambootstrap_seed_val This random number is provided to assure that there is comparability between runs. bootstrap_seed Conduct rapid bootstrapping? (-x) 1 perl !$compute_mr && !$thorough_opt && !$ancestral_states && !$fast_tree && !$mulparambootstrap_seed && !$startingtreeonly && !$compute_ml_distances && !$bipartitions && !$log_likelihood && !$compute_mr && !$specify_runs This option offers a novel rapid Bootstrapping algorithm that is faster by at least one order of magnitude than all other current implementations (RAxML 2.2.3, GARLI, PHYML). The results obtained are qualitatively comparable to those obtained via the standard RAxML BS algorithm and, more importantly, the deviations in support values between the rapid and the standard RAxML BS algorithm are smaller than those induced by using a different search strategy, e.g. GARLI or PHYML. This rapid BS search can be combined with a rapid ML search on the original alignment and thus allows users to conduct a full ML analysis within one single program run. bootstrap_seed_val Enter a random seed value for rapid bootstrapping perl ($value) ? " -x $value" : "" 12345 2 perl $bootstrap_seed && !$mulparambootstrap_seed && !$startingtreeonly && !$compute_ml_distances Please enter a random seed for the -x option (eg 12345) perl $bootstrap_seed && !defined $bootstrap_seed_val This random number is provided to assure that there is comparability between runs. mlsearch Conduct a rapid Bootstrap analysis and search for the best-scoring ML tree in one single program run. (-f a) perl ($value)?" -f a ":"" 1 2 Conduct a Rapid Bootstrap analysis (-x) and search for the best-scoring ML tree in one single program run. perl !$compute_mr && !$thorough_opt && !$ancestral_states && !$fast_tree && !$bipartitions && !$startingtreeonly && $bootstrap_seed && !$compute_ml_distances printbrlength Print branch lengths (-k) perl ($value)?" -k":"" 0 2 The -k option causes bootstrapped trees to be printed with branch lengths. The bootstraps will require a bit longer to run under this option because model parameters will be optimized at the end of each run under GAMMA or GAMMA+P-Invar respectively. specify_bootstraps Specify an Explicit Number of Bootstraps 1 perl ($bootstrap_seed || $mulparambootstrap_seed) && (!defined $use_bootstopping || !$use_bootstopping) bootstop Bootstrap iterations (-#|-N) perl ($bootstrap_seed || $mulparambootstrap_seed) && !$use_bootstopping && !$startingtreeonly perl " -N $value" 100 2 Please enter number of bootstrap reps desired (eg 100) perl $specify_bootstraps && !defined $bootstop Sorry, the value of bootstraps cannot exceed 1,000 perl $bootstop > 1000 Specifies the number of alternative runs on distinct starting trees. If 10, RAxML computes 10 distinct ML trees starting from 10 distinct randomized maximum parsimony starting trees. In combination with the Random seed for rapid bootstrap (-x) invoke a rapid BS analysis. use_bootstopping Let RAxML halt bootstrapping automatically perl ($bootstrap_seed || $mulparambootstrap_seed) && (!defined $specify_bootstraps || !$specify_bootstraps) 0 Please select to "specify bootstraps explicitly" or "automatically halt bootstrapping" perl ($bootstrap_seed || $mulparambootstrap_seed) && (!$use_bootstopping && !$specify_bootstraps) 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. freq_bootstopping Stop Bootstrapping Automatically with Frequency Criterion perl $use_bootstopping && !$mr_bootstopping perl ($value) ? "-N autoFC":"" 0 Please choose either majority rule or frequency criterion perl $use_bootstopping && ( !$mr_bootstopping && !$freq_bootstopping ) If you want to use the bootstopping criteria specify "-# autoMR" or "-# autoMRE" or "-# autoMRE_IGN" for the majority-rule tree based criteria (see -I option) or "-# autoFC" for the frequency-based criterion. Bootstopping will only work in combination with "-x" or "-b" mr_bootstopping Stop Bootstrapping Automatically with Majority Rule Criterion (recommended) perl $use_bootstopping && !$freq_bootstopping 1 Please choose either majority rule or frequency criterion perl $use_bootstopping && (!$mr_bootstopping && !$freq_bootstopping) If you want to use the bootstopping criteria specify "-# autoMR" or "-# autoMRE" or "-# autoMRE_IGN" for the majority-rule tree based criteria (see -I option) or "-# autoFC" for the frequency-based criterion. Bootstopping will only work in combination with "-x" or "-b" mrbootstopping_type Select Majority Rule Criterion: (autoMRE is recommended) perl $mr_bootstopping perl "-N $value" autoMR autoMRE autoMRE_IGN autoMRE Please choose a Majority Rule criterion perl $mr_bootstopping && !defined $mrbootstopping_type use_apobootstopping Use a posteriori bootstrapping 0 perl !$use_bootstopping Sorry, you cannot use a posteriori bootstrapping with the -b or -x options perl $use_apobootstopping && ($bootstrap_seed || $mulparambootstrap_seed) This option is used when a bootstrap analysis is already completed. Upload an input file, a tree, and bootstrapped trees. You must upload a set of bootstrapped trees uploaded as a single file using the -z option. Once this selection is made, the user must select whether the bootstopping threshold is determined using frequency criteria or majority rule criteria aposterior_bootstopping Select the criterion for a posteriori bootstrapping analysis perl $use_apobootstopping perl "-I $value" autoFC autoMR autoMRE autoMRE_IGN In order to use the a posteriori bootstrapping option, you must upload a set of bootstrapped trees below perl !defined $apo_tops 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" apo_tops File with topologies for a posteriori bootstopping (-z) perl " -z apotopologies_file.tre" 2 apotopologies_file.tre perl $use_apobootstopping && !defined $bunchotops all_outputfiles *