PAGAN2

PAGAN2 is a general-purpose method for the alignment of DNA, codon and amino-acid sequences as graphs. It aligns sequences either with pileup or, when related by a tree, using phylogeny-aware progressive alignment algorithm. In both cases it uses graphs to describe the uncertainty in the presence of characters at certain sequence positions. PAGAN2 is largely compatible with PAGAN but implements new algorithms for alignment anchoring and memory handling. As a result, PAGAN2 can align sequences of several hundreds of kilobases in length.

If you use Linux and just want to use PAGAN2 (i.e. not see the source code and compile it from scratch), please download the latest bundle (pagan2.linux64.[latest].tgz) from the bin folder. This package includes all helper applications needed e.g. for the guidetree inference.

See the original PAGAN documentation for the usage instructions.

---

Compilation of PAGAN2

PAGAN2 uses the NCBI TOOLKIT and Boost libraries. The NCBI TOOLKIT does not compile with GCC compiler newer than 4.8 and also requires a rather old version of libc. The code compiles well on Ubuntu 14.04 and instructions are provided for building it with that using Docker.

If that doesn't work on your system, you can try to use PAGAN2 inside the precompiled Docker container available at Docker hub. Please read the instructions for that below. If your system supports Singularity but no Docker, you can easily convert the PAGAN2 Docker container to run under Singularity. Please read the instructions for that at the bottom.

Instructions

Both options naturally require that one has Docker installed...

1. Using the precompiled Docker image from Docker hub

Get the image and rename it as pagan2

docker pull ariloytynoja/pagan2
docker tag ariloytynoja/pagan2 pagan2

Get some data

wget https://raw.githubusercontent.com/ariloytynoja/pagan-msa/master/examples/454_pileup/454_reads.fas 

Run PAGAN2 inside the container

docker run --rm -v `pwd`:/data \
  pagan2 --pileup --homopolymer \
  -q 454_reads.fas -o 454_aligned

Here Docker the options tell to remove the container after finishing (--rm) and map the file systems (-v `pwd`:/data). As the options are always the same, we can write them in a script file and call that instead.

Create a script file to simplify the command

cat > pagan2.sh << EOF 
#!/bin/bash
docker run --rm -v `pwd`:/data pagan2 "\$@"
EOF

chmod +x pagan2.sh 

./pagan2.sh --pileup --homopolymer \
  -q 454_reads.fas -o 454_aligned

2. Compiling PAGAN2 from scratch using Docker

Get the dockerfiles

 git clone https://github.com/ariloytynoja/pagan2-docker-build.git
 cd pagan2-docker-build/

Build the libs and the app

docker build \
    -t pagan2libs \
    -f Dockerfile.pagan2libs \
    .

docker build \
    -t pagan2app \
    -f Dockerfile.pagan2app \
    .

Option 1: Get the statically linked binary out

docker create \
    --name pagan2app \
    pagan2app

docker cp pagan2app:/pagan2/src/pagan2 ./pagan2

The binary pagan2 contains all dependencies and can be used on all modern Linux systems (Ubuntu 14.04 and later).

Option 2: Build the helper apps and get them out

docker build \
    -t pagan2app \
    -f Dockerfile.pagan2apps \
   .
   
docker create \
    --name pagan2apps \
    pagan2apps

docker cp pagan2apps:/pagan2/pagan2apps.tgz ./pagan2apps.tgz

tar xzf pagan2apps.tgz

./pagan2/bin/pagan2 --pileup --homopolymer \
  -q 454_reads.fas -o 454_aligned

Option 3: Create a small PAGAN2 image and run it as above

docker build \
   -t pagan2 \
   -f Dockerfile.pagan2 \
   .
   
docker run --rm -v `pwd`:/data \
 pagan2 --pileup --homopolymer \
 -q 454_reads.fas -o 454_aligned

The Docker container should run on non-Linux systems.

3. Using the precompiled Docker image with Singularity

Get the image

singularity build pagan2.sif docker://ariloytynoja/pagan2

Get some data

mkdir data
wget https://raw.githubusercontent.com/ariloytynoja/pagan-msa/master/examples/454_pileup/454_reads.fas -P data/

Run PAGAN2 inside the container

singularity run -B `pwd`/data:/data \
  pagan2.sif --pileup --homopolymer \
  -q data/454_reads.fas -o data/454_aligned.fas

Note that we downloaded the data to a directory called data and run the analysis outside of that directory. The reason for this is that Singularity doesn't support the Docker WORKDIR and we have to specify the path with "data/" prefix. This prefix can be added to any path but with this arrangement it is intuitive. As the options are always the same, we can write them in a script file and call that instead.

Create a script file to simplify the command

cat > pagan2.sh << EOF 
#!/bin/bash
singularity run -B `pwd`/data:/data pagan2.sif "\$@"
EOF

chmod +x pagan2.sh 

./pagan2.sh --pileup --homopolymer \
  -q data/454_reads.fas -o data/454_aligned