Gene Conservation Laboratory
Statistics Program for Analyzing Mixtures (SPAM) Software
SPAM - Genetic Stock Identification Software
- SPAM Homepage
- SPAM FAQs
- Genetic Analysis Links
- Version History
SPAM - What is it?
Statistics Program for Analyzing Mixtures (SPAM) estimates the relative contributions of discrete populations to a mixture sample, solving what is commonly referred to in fisheries as the mixed stock analysis (MSA) or genetic stock identification (GSI) problem. The software was mainly developed here at the Gene Conservation Lab and is available for free.
What can you do with it?
See the Version 3.7b documentation for a full description. For additional reference, see a version history (debug and enhancement history).
SPAM can run in either Estimation or Simulation mode. In estimation mode, SPAM derives mixture proportion estimates by numerically solving a conditional maximum likelihood problem. The likelihood is conditional on the assumption that the distributions of identifying characteristics are known without error for each population potentially contributing to the mixture. In simulation mode, SPAM samples from a known mixture of baseline populations, then estimates the mixture proportions. This allows one to investigate mixture sample sizes, 'identifiability' of potential populations or population assemblages, etc.
Besides proportion point estimates, SPAM 3.7b can be used to generate confidence intervals, test equality of mixtures that have been independently sampled, conduct likelihood ratio tests to compare competing mixture models, assess potential contribution estimate bias arising from using very large baselines, and investigate changing detection power as a function of mixture sample size.
Mixture Analysis Papers and Posters
See papers, drafts and posters related to the mixture analysis and SPAM. You can also submit your own paper or a preprint to be posted on this website.
What goes well with it? (recommended sites)
Collection of genetic data analysis links.
Questions, Comments, Suggestions, Problems?
See the SPAM FAQ page. If you don't find what you need there, send us an e-mail.