Banana Slug Genomics

* attempted to run pre-compiled binaries, but they expect non existent shared libraries in /usr/lib.

Running make gives the following error: g_band.c:44: error: static declaration of 'g' follows non-static declaration inc/extvab.h:20: error: previous declaration of 'g' was here g_band.c:44: error: static declaration of 'h' follows non-static declaration inc/extvab.h:20: error: previous declaration of 'h' was here make: *** [g_band.o] Error 1

Googling this problem reveals that one way to fix this problem is take the static definition off of g_band.c:44. http://www.mail-archive.com/debian-med-commit@lists.alioth.debian.org/msg00746.html

 cp g_band.c g_band_orig.c

Changed line 44 to now read:

 //static int g, h, m;                           /* g = G, h = H, m = g+h */
 int g, h;                               /* g = G, h = H, m = g+h */
 static int m;

Now build finishes fine other than the following warning:

 make: warning:  Clock skew detected.  Your build may be incomplete.

For installation I made the directory

/campusdata/BME235/bin/euler 

and symlinked the executables and scripts into that directory.

Add the following to your path:

/campusdata/BME235/bin/euler

set the following environmental variable:

EULERBIN=/campusdata/BME235/bin/euler

Make sure that your LD_LIBRARY_PATH and LD_RUN_PATH variables do NOT point to our /campusdata/BME235/lib directory. For some reason this seems to get in the way of some of the standard libraries used by the 4.1 compiler. Since this program was compiled against 4.1 this environmental variable needs to stay default. * I think the system has a default LD_LIBRARY_PATH variable that works fine, so an unset is not appropriate in this case without re-setting the default minus our /campusdata/BME235/lib/../lib64 path.

I am currently doing an assembly on one of the test datasets provided on the euler home page. To see this dataset and the script used to run it go to:

/campusdata/BME235/programs/euler/testData

Once I can get this one test dataset working I won't bother with the other they provided, and move on to testing on Pog. My main goals in running the program on this test data are to eliminate as many confounding variables in getting proper execution, and also to explore and better understand the format that euler expects. Here is what I have discovered about Euler's required data format so far:

• All reads need to be in a single fasta file, however this file may contain reads from different runs/technologies.
• The name of the read matters and is used by Euler to determine whether each sequence has a mate pair, what the mate pair is, and also the distance range of that mate to its pair. This is how you differentiate between reads from different technologies or library protocols.
  • How Fasta id lines are deciphered is defined by the user in a file called names.rul and this file must be placed in the directory you are running your assembly from (your cwd when you start execution).