|
NEW
- SPAM
version 3.7b
- Mixture Analysis Papers and Posters
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.
Look here for 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
Here you can find 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. |
