Troubleshooting

For issues with compiling or getting Transposome installed, see the suggestions below. If there are other issues not listed here, please send me an email or document the issue on the issue tracker page.

---

Troubleshooting issues compiling

On some Linux distributions, you may see an error message when you try to run perl Makefile.PL. Here are some suggestions for resolving these issues. If you see a message like:

[ERROR]: Encountered problems trying to compile source. Contact the author if you are unable to resolve this issue.

Here are the errors: main_community.cpp:112:29: error: ‘getpid’ was not declared in this scope. Exiting.

Try editing the file src/main_commmunity.cpp and add "#include <unistd.h>" before the other include lines. Then, just run perl Makefile.PL again see if that solves the issue.

On some Macs, you may also encounter some issues depending on your compiler. If you encounter this message after running perl Makefile.PL:

[ERROR]: Encountered problems trying to compile source. Contact the author if you are unable to resolve this issue.

Here are the errors: graph_binary.h:23:20: error: malloc.h: No such file or directory. Exiting.

Edit the file src/graph_binary.h by commenting out the line "#include <malloc.h>" line (i.e., change the line "#include <malloc.h>" to "//#include <malloc.h>".

Troubleshooting out of memory issues

If you attempt to analyze a very large number of sequences, it is possible that the clustering methods could consume all your computer's memory. The first thing to note is that it isn't necessary to analyze all of your reads to get an accurate result, and adding more data likely will not change the result. Here are a few suggestions for getting around these issues.

1. In your configuration file, set the in_memory attribute to "0" for better memory conservation.
2. Downsample your data. There are code examples listed in the section on choosing the appropriate genome coverage and in the Tutorial section for sampling your sequence data.

Troubleshooting issues installing Perl dependencies

If running perl Makefile.PL generates no errors, then it is okay to proceed with building the package. If there are warnings about missing dependencies, you will need to install those first. Installing all the dependencies can be handled most easily with cpanminus. For example,

git clone https://github.com/sestaton/Transposome.git
cd Transposome
cpanm --installdeps .
perl Makefile.PL
make
make test
make install

Alternatively, you can use the CPAN shell that is shipped with Perl to install the dependencies.

git clone https://github.com/sestaton/Transposome.git
cd Transposome
perl Makefile.PL
## If you see a specific module is not installed, for example, DateTime,
## you can install it with the following command.
## If there are multiple missing modules, just include them all in the same command,
## for convenience.
##
## Note that the following command might need to be preceded by "sudo" if you
## are using system Perl.
perl -MCPAN -e 'install DateTime'
perl Makefile.PL
make
make test
make install

Running these commands should install Transposome without warnings, and you can test the installation by typing transposome --help.