Step 0: Clone this repository.
git clone https://github.com/ekmolloy/fastmulrfs.git
cd example
Step 1: Preprocess gene family trees. This transforms each MUL-tree into singly-labeled (and potentially unresolved) tree. Importantly, the optimal solution to the MulRF supertree optimization problem for the MUL-trees is the optimal solution to the RF supertree optimization problem for the preprocessed MUL-trees (Theorem 2). NOTE: If you haven't already, first install TreeSwift.
python ../python-tools/preprocess_multrees_v3.py \
-i g_trees-mult.trees \
-o g_trees-mult-for-fastrfs.trees \
--verbose
Step 2: Run FastRFS, a heuristic for estimating a RF supertree, on the preprocessed gene family trees. NOTE: If you haven't already, first install FastRFS, as described here.
../external/FastRFS/build/FastRFS \
-i g_trees-mult-for-fastrfs.trees \
-o fastmulrfs.tree &> fastmulrfs.log
Also see this bash script.