mothur v.1.7.0

We hare happy to announce the release of mothur v.1.7.0 in time for the US Thanksgiving Holiday. So as you gather around the table and tell your family what you are thankful for, be sure to thank your mothur. This has been a great month for mothur. The manuscript describing mothur has been published in Applied & Environmental Microbiology. Also, the manuscript describing the aligner that is used in mothur has been accepted by PLoS ONE and will hopefully be posted soon. I give you full permission to cite both liberally. Because of this we have had a fair amount of traffic by people downloading mothur - so if this is your first update, welcome. Many of you may remember that last month we launched two new ventures a facebook page and a user forum. We are up to 53 “friends” and have gotten a steady stream of forum users. Please feel free to use the wiki, forum, and facebook as means of communicating with and helping each other. December also promises to be a promising month as The Mothur, is due to release dotur v.2.0 “Ruth” on December 6th.

This update has a number of features that people are sure to find useful. First, we have added several commands for processing and characterizing OTUs. Second, we have added an alternative algorithm for clustering sequences using the furthest neighbor algorithm, which is based on the algorithm described by Sun and colleagues (2009). We are unable to replicate the numbers they describe for mothur, so take that part of their analysis with a grain of salt. What we have found is that the memory requirements using the hcluster command are minimal; however it tends to run a bit slower than the cluster command. Finally and most significantly, we have added two methods for classifying sequences - the k-Nearest Neighbor algorithm and RDP’s Bayesian Classifier. These data can then be used to assign sequences to phylotypes. As with everything we do, our goal is to provide the flexibility to use any classifier you want with any taxonomy scheme you want. With these additions, we now have the ability to pursue OTU, phylotype, and hypothesis testing-based approaches. We will continue to add features, including those that you suggest, with the goal of making mothur everything you need it to be.

New commands

  • get.otulist command - Outputs a .otu file for each distance specified containing the bin number and a list of the sequences in that bin.
  • hcluster command - clusters distance files of any size. It is slower than the cluster command, which stores the distance in RAM to increase performance, so if you can fit your file in RAM use cluster.
  • classify.seqs command - classifies sequences using the knn or bayesian methods.
  • phylotype command - creates .list, .rabund and .sabund files from a taxonomy file.

Updates

  • read.otu command - Added a groups parameter to specify which groups you would like to use.
  • summary.shared and collect.shared commands - added an all option, which estimates the combined richness of all your groups when using the sharedchao and sharedsobs calculators.
  • trim.seqs command - Added a qtrim parameter, which gives an additional method of processing quality data.
  • libshuff command - Fixed a bug, which could affect P-values.