Computes dinucleotide (AA/TA/TT) periodicity, elastic energy to form a nucleosome, and a prediction of nucleosome occupancy/free energy based on sequence periodicity

For the analysis of multiple short sequences (~200 bp). Expects fasta file with one or more sequences, computes averaged periodicity of AA/TA/TT, or the minimum elastic energy of all possible nucleosomes that could be formed.

Based on MD-averaged helical parameters force-constants and averaged values, plus reference values of helical values in "canonical nucleosomes". Requires database files (stif_bsc1_k_avg_miniabc_dinuc.dat, refnuc_bp.dat, and stif_bsc1_k_avg_miniabc.dat), which were sent to me by F. Battistini, from the Orozco group.

Based on "Sequence-based prediction of single nucleosome positioning and genome-wide nucleosome occupancy", van der Heijden et al. DOI: 10.1073/pnas.1205659109

The code is basically a 1-to-1 translation of a python script that was sent to me by J. van Noort, without any mention of license. There are four parameters in the model. Two have been hardcoded as #defines, the other two are defaulted to the recommended values but can be changed.

1. Seqan 2.2.0 (other versions untested, but probably fine)
2. Eigen 3.3.0 (other versions untested, but probably fine)
3. OpenMP (but could be removed with loss of functionality)

[email protected]:guillemportella/periodic_elastic.git
cd period_elastic

Fix the CMakeLists depending on your library location.

In clust1-headnode-1 this shoudl work

module load cmake/3.7.1 gcc/6.2.0 BOOST/1.56.0
cmake3 . -DCMAKE_C_COMPILER=/home/portel01/programs/compilers/gcc-6.3/bin/gcc -DCMAKE_CXX_COMPILER=/home/portel01/programs/compilers/gcc-6.3/bin/g++