This container contains the application for analyzing and scoring gene families for over-clustering using a phylogeny based approach. Each given family fasta is search against a database of outgroup sequences to find the closest outgroup sequences. The family sequences along with the closest outgroup sequences are used to build a Maximum-Likelihood (ML) phylogeny and rooted using the monophyletic outgroup clade or the closest outgroup. The family phylogenies are then scored using the proportion of ingroup sequence pairs that appear to diverge after one or more outgroup sequences. Finally, monophyletic ingroup clades are extracted from the family phylogenies based on predefined species representation cutoff.
docker pull akshayayadav/overcl-detection-correction
Create a directory <my_directory> containing directory named family_fasta, a fasta file named outgroup-sequences.fa containing all the outgroup sequences and species label file named species.profile containing species labels for the ingroup and outgroup species. The species.profile contains 2 columns: first column contains the names of all ingroup and outgroup species and second column containing '1' or '0' labels for ingroup and outgroup names. The names in the species.profile file MUST be present as prefixes of sequence headers in the family fasta files and the outgroup-sequences.fa fasta. Also, the sequence of outgroup names in the species.profile file should be same as the relative distance from the ingroup species. For example, if outgroup1 is closer to ingroups than outgroup2 that outgroup1 should be mentioned before outgroup2 in the species.profile file.
docker run -v <absolute_path_to_data_directory>:/data akshayayadav/overcl-detection-correction run_analysis.sh