It would be helpful to have more details as to what you mean by STRUCTURE(?) didn't work. Did the program not run properly or do you simply disagree with the result that was produced?

If it is a problem with the program not running properly, try looking on the google groups forum for STRUCTURE that is located here: https://groups.google.com/forum/#!forum/structure-software

If you are having trouble interpreting results, you could try using STRUCTURE HARVESTER to get some guidance on the interpretation of the results. (http://taylor0.biology.ucla.edu/structureHarvester/)

If you disagree with the results, try taking a look at TESS (http://membres-timc.imag.fr/Olivier.Francois/tess.html) and seeing if TESS produces similar results.

You could also look at some of the functions included in the ADEGENET package in R (http://adegenet.r-forge.r-project.org/)

Also, take a look at this paper to see some other options:

Bayesian clustering algorithms ascertaining spatial population structure: a new computer program and a comparison study

CHIBIAO CHEN1,2, ERIC DURAND2, FLORENCE FORBES1 andOLIVIER FRANÇOIS2

Article first published online: 6 APR 2007

DOI: 10.1111/j.1471-8286.2007.01769.x

thanks. The program can work, but every time, there are no files in the result folder. I tried in win7, win8 and winxp, all are the same. I did not know why, So I want to use another one to do this work. I will try these program.

Thanks again

It may be that you haven't set something up properly in your run. Try going to the google groups forum for STRUCTURE and asking there for some help or search the existing questions as I know that other people have reported that problem and they've been able to solve it fairly easily.

I contacted the author of structure, but he has no way to resolve this.

You could try BAPS or GENELAND

The main Bayesian programs besides Structure are BAPS, Geneland, and TESS. Most journals that I have submitted to always recommend two programs to verify assignments. I typically run one set with non-spatial coordinates (Structure) and one with spatial coordinates (BAPS). Structure's non-spatial model is the best in my experience and the other three perform best with spatial information (both Geneland and TESS require spatial coordinates to perform best).

BAPS's spatial models are really great, but the non-spatial clustering of individuals or groups tends to give a high number of false clusters if you have strong isolation-by-distance. Geneland and TESS also suffer from this problem, but even more so.

So, for Structure, the MCMC runs are on the output, it just doesn't create a folder? I have Structure running on both my Mac and PCs (Win 7 and 8), and have not had that problem. It may be a problem with your computer instead of the program. I have experienced several bugs with Structure, but usually it fails to runs when I try to execute a job. That bug requires you to shut down structure and reload the job. Good luck, if you can only run a single Bayesian program, I would back it up with something like a PCA that shows concordant population structure.

Good luck.

With STRUCTURE, I have frequently noticed that same problem in Windows.

The software can run the analysis properly but is not able to write the results because the file path is too long (too many folders into folders into folders...).

I suggest you try again to run your analysis with a short file path such as:

C:\STRUCTURE\analysis1

Let me know if it solves your problem.

cheers

Marc

I will try it following your suggestion.

Hi, Marc

I tried again after your suggestion. But it still no files in the result folder. when it runs, all "Est Ln P(D)" are zero.

I had a similar problem. The output files were not being written to the designated output folder where I expected them. I ran a search using the file name I expected, and found the output files in temporary items. I could then copy them across to where I wanted them. Not an ideal solution, but it worked. Maybe you have the same problem.

I think it might have been something to do with Structure being installed on a part of the computer that required administrator permissions to write files, but I'm not sure. Someone with more knowledge of the Windows operating system may be able to tell you the reason for the problem.

I remember having a similar problem with the administer permissions on my PC at work. If you do have administrator access, right click on Structure's icon and choose "run as administrator". I remember that worked for some of my installed programs (I think DistLM, not sure). If you do not have administrator access, then find the person who is or get your IT person to grant you access.

I hope this helps,

Liz

[email protected]

I had similar problem.