FAMD

DNA fingerprinting (AFLP, RAPD, ISSRs, etc.) studies will inherently bring uncertainties with them, whether that be noise associated with the technique, or errors made in scoring, etc..

You can score your data using absence/presence (0/1). This means that you will attempt to partition data points about which you are unsure into absences and presences (to be conservative, probably absences rather than presences).

Alternatively, you can allow for uncertainties: (0/?/1). This enables you to later have a second look at your data and ask what impact this ambiguous or missing data may have on your analysis.

FAMD is a little, easy to use application specially designed to help you doing that. Specifically addressing missing (or ambiguous) data, FAMD implement the following:

· minimum/ maximum/ average (dis)similarity calculation allowing the estimation of the extent to which missing data may impact on the analysis. · Jaccard's, Dice, SMC similarities, NeiLi and Euclidean distances · output of distance matrices · missing data resampling · estimation of Shannon's index by bootstrapping. · UPGMA, NJ, strict and majority rule consensus trees · estimation of R-support · PCoA (principal coordinate analysis) and 3D viewer with bitmap and metafile support · AMOVA for all implemented (dis)similarity measures · export of data to a number of formats such as Nexus, Arlequin project, hindex, Hickory (Nexus Alleles block), GenePop, NTSYSpc, Structure, hindex. · Bayesian estimation of population null allele frequency and calculation of inter-population distances.