ALFRED : Distance Estimator for Phylogenetic Inference

ALFRED is an Alignment Free Distance Estimator software for Phylogenetic Inference. It takes as input a set of n sequences and ouputs an n x n matrix of distance estimate for use in Phylogenetic inference.


  1. ALFRED source code
  2. Datasets

Installation and Usage

External Dependencies that need to be installed

  • A modern, C++11 ready compiler such as g++ version 4.8 or higher or clang version 3.2 or higher.
  • The cmake build system (Version >= 2.8.11).
  • A 64-bit operating system. Either Mac OS X or Linux are currently supported.
  • Git version control system

Dependices made available with the gz file


Download the alfred.tar.gz file available here. Extract the contents as follows:

tar xvzf alfred.tar.gz

Next, create a build directory. For example,

 mkdir build
 cd build

Finally, configure and build the executable 'alfred.x'. Continuing the example,

 cmake ..

If successfully built, alfred.x should be available in the build directory.


Preparing Input

Place the input sequences into a single fasta files. Make sure that the header of the fasta file has the appropriate sequence name. For example, PHYLIP doesn't accept sequence names longer than 10 characters. Also, make sure that the input file has one of 'fasta' or 'fa' or 'fas' file extension. Note that the program accepts only sequences with valid DNA or protien alphabets. An example is as shown below.



Compute greedy alignment free estimator

Run the program by providing the pepared input file with -f option, the output file with the -o option, and the number of mismatches to allow with -x option. An example is shown below:

build/alfred.x -f primates.full.fas -o alfred.primates.x8.out -x 8

Output is generated in a matrix format, which can be fed directly to PHYLIP


Datasets used in our experiments are available from here. The contents of the files are as follows:

aliases.xlsx contains the alises we used for the organism names so as to construct the tree.

data sub-directory contains all the input data files and the reference trees. The input files are in fasta format and has .fa extension and the refrence trees are in phylip format and has .tree extension. For roseobacter dataset, multiple sequence alignment is also given. MSA for BaliBASE datasets can be downloaded from the BaliBASE website.

runs sub-directory contains all the output matrix files and the trees with the best RF.dist score. Ouput matrix files use the format as accepted by PHYLIP.


  1. Thankachan, S. V., Chockalingam, S. P., Liu, Y., Krishnan, A., & Aluru, S. A greedy alignment-free distance estimator for phylogenetic inference. In Computational Advances in Bio and Medical Sciences (ICCABS), 2015 IEEE 5th International Conference on (pp. 1-1). IEEE. 2015.

  2. Thankachan, S.V., Chockalingam, S.P., Liu, Y., Apostolico, A. and Aluru, S., 2015. ALFRED: a practical method for alignment-free distance computation. Journal of Computational Biology.

alfred.tar.gz1.28 MB