Having trouble running your analysis?


  1. Your files need to have UNIX line breaks. Even if you think you have done this, check again! You can do this with a text editor such as TextWrangeler.

  2. Many of the programs on the cluster (MrBayes, PAUP) are a bit finicky about the files they accept as input. A few file formatting tips are provided below.

    • Use only alphanumeric characters in your taxon names. DO NOT use spaces or dashes. An underscore can be substituted for a space or dash.

    • MrBayes will only recognize a subset of the characters that PAUP recognizes. See page 10 of the MrBayes 3.1 manual.

    • Be sure that you designate any special characters in the Nexus datablock (e.g., "missingdata=?", "gap=-").

  3. It is not uncommon for a node on the cluster to malfunction. When this happens it is often an entire computer that malfunctions and two processors are affected. If your jobs are failing in multiples of two, check to see if these jobs are using the same computer. If so, it could be the computer that is causing your problems.

  4. Do not use spaces in your file and directory names

  5. Linux is case sensitive. When the instuctions specify that a file or directory have a certain name, make sure that you type the name exactly as given.

  6. Does your MrBayes input file have a ".nexus" extension?

  7. When running multiple PAUP, MrBayes, Modeltest, MrModeltest, and PAML analyses, run each one from its own folder. This does not apply to the individual replicates of PAUP bootstrap analyses.

  8. Occasionally a program will not run on the cluster due software issues on the cluster. If you think this is occurring, send an email to the cluster managaer Kelly Crone Willis (crone@umsl.edu). Be sure to include the name of the program you were using and the .out and .err files that were produced.