TreeOTU TreeOTU: Operational Taxonomic Unit Classification Based on Phylogenetic Trees Dongying Wu, Ladan Doroud, Jonathan A. Eisen University of California, Davis, Davis, California 95616, USA TreeOTU is a package of perl scripts that classify Operational Taxonomic Unit (OTU) based on phylogenetic trees. The package including the following components: 1. Main script: TreeOTU.pl TreeOTU.pl takes a rooted phylogenetic tree and a PN (Position of Node) cutoff (0<=PN<1) and outputs a set of OTUs Usage: TreeOTU.pl -i input_tree -c PN_cutoff -o output_file Input format: -i inputtree, the input tree must be in newick format -c PN cutoff, PN cutoff must >=0 and <1 Output format: -o output file, if no output file is define, the script outputs to STDOUT The format of the output is demonstrated by the following example: input_tree_PN_cutoff_0.05 A,B C D,E The first column is the OTU set description including tree file name and PN cutoff, from the second column to the end of the line are the OTUs. different OTUs are separated by spaces while the taxa in each OTU are separated by commas. 2. OTU comparison scripts A. Compare OTU sets based on Adjusted Mutual Information (AMI): OTUcompare_AMI.pl OTUcompare_AMI.pl compares one of OTU set (Query) with multiple sets of OTUs (Target), and outputs the AMI values (Ajusted Mutual Information) between the query and target(s). AMI is a number between 0 and 1. AMI value of 1 indicates two OTU sets are identical, the more different the two OTU sets, the smaller AMI gets. Usage: OTUcompare_AMI.pl -query query_OTUset -target target_OTUset -shuffle 100 Input format: -query query_OTUset, this input is a file with only ONE line of OTU set, the format of the input is exactly the same as the output of TreeOTU.pl -target target_OTUset, this input target file can contain more than one OTU sets. One OTU set takes one line, the format of the OTU layout is exactly the same as the output of TreeOTU.pl -shuffle an_integer, the number of shuffle times to determine the expect value in AMI calculation, the default value is 100 Output format: The script outputs results to STDOUT, the output format is examplified by the following line: AMI=0.5696 Query_ID=BA00028_0.080 Target_ID=concat38_0.010 Shuffle=100 Query_taxa_count=4573 Target_taxa_count=4436 Shared_taxa_count=4424 The output explains itself. Only the shared taxa in both OTU sets are included in the AMI calculation. B. Identify the maximum F1 score of one OTU in a OTU set: OTUcompare_Fmeasure.pl This script compares one OTU (query) against all the OTUs in a OTU dataset (target), and report the maximum F measure score. Usage: OTUcompare_Fmeasure.pl -query query_OTU -target target_OTUsets -shuffle 10 -beta 1 Input format: -query query_OTU, the OTU group of interest, format: A,B,C,D -target target_OTUsets, the target OTU set that the query is compared against. The input target file can contain more than one OTU sets. One OTU set takes one line and the format of the OTU layout is exactly the same as the output of TreeOTU.pl -shuffle integer, target random shuffling times to evaluate the expected maximum F measure score (the default is 10) -beta: F measure beta factor, the default is 1 Output format: The script outputs results to STDOUT, and the output takes the format of the following line: targetID=concat38_0.600 Adjusted_max_Fscore=0.0032 Max_Fscore=0.0640 Query_taxa_count=135 Query_taxa_evaluated=132 BETA=1 Shuffle=5 Expected_maxF=0.0610 3. Tree rooting scripts A. Mid-point rooting: MidpointRooting.pl Usage: MidpointRotting.pl -i input_tree -o output_tree (input must be in newick format) B. Rooting a tree between two nodes defined by six taxa: SixPointRooting.pl The script roots the tree in the middle of two nodes. Three taxa define one node, thus the script needs 6 taxa to carry out the rooting process. Usage: SixPointRooting.pl -i input_tree -l taxa_list -o output_tree input format: -i input tree, the tree must be in newick format -l input taxa list, two lines, each line includes 3 taxa separated by spaces C. Rooting the tree according to an input outgroup: OutgroupRooting.pl The script takes a list of taxa, and tries to find the best matching monophyletic clade as the outgroup, evantually roots the tree in the middle of the edge that connects the outgroup and the rest. Usage: OutgroupRooting.pl -i input_tree -l input_taxa -o output_tree input format: -i input tree, the tree must be in newick format -l input list of taxa as the outgroup, can be non-monophyletic in the tree. The format of this input is multiple lines with one aceesion in each line. output format: -o the rooted tree. There is a log file with the extension ".OutgroupRootingLog" that documents the actual outgroup the script decides to use, and its similarity with the input outgroup (F1 score, 1 means identical, the F1 score should be >=0.5 for the script to carry out the rooting) 4. Additional Data The file TreeOTU_ali_tree_taxa.tgz inlcudes alignments, trees and taxonomy information for 40 PhyEco marker families and ssu-rRNAs from the IMG database as well as the ssu-rRNAs from the "All-Species Living Tree" project.

Dongying Wu; Ladan Doroud; Jonathan A. Eisen