Multi locus approach

I am working on land plants, and have just started my barcode activity.
As by the recent publication, the multi locus approach for developing
the barcode seems to be the most accepted procedure (though debatable).
My query is: If i have sequences from two and more locus (which are
probable barcode locus) and I aim at getting a evolutionary or/and
phylogenetic relationship between species based on a combination of
these locus, what should be the methodolgy to do so. Should the distance
be computed for the different locus and then added and normalized and
then using UPGMA or NJ one can have the tree. Or there exists a well
established protocol for this acivity. Kindly help me as I am naive to
this.