Programs

The eThread software distribution includes the following main programs: econtact, eextract, emodel, erank-modeller, erank-tasser and ethread. To get the list of available options for each program, simply execute it with no arguments. Below is the full description of the programs, options and formats.

ethread

ethread selects structural and functional templates using machine learning on results reported by individual threading algorithms. It also carries out an alignment refinement procedure to construct a set of consensus threading alignments.

Mandatory arguments:

-i input_file, where input_file is a target sequence in FASTA format.

-o output_file, where output_file will contain consensus alignments constructed for your target sequence.

-l lib_file, where lib_file is the entire threading library in FASTA format.

Optional arguments:

-d lib_type, where lib_type is the threading library type and should be either chain or domain. The default type is chain.

-b seq_cut, where seq_cut is a maximum sequence identity of selected templates to the target. This argument should be used only for benchmarks. The default value is set to 1.0, which means that all identified templates will be used in modeling.

-s svm_cut, where svm_cut is a threashold used for template selection by machine learning. The default value is 0.15, which was optimized during extensive benchmarks.

-m ali_min, where ali_min is a minimum length of alignment. The default value is 20, which means that only alignments that have at least 20 aligned residues will be reported.

-f function_file, where function_file will contain probabilities for all identified structural templates to be also functional templates. These probabilities are calculated separately for various aspects of molecular function including ligand-, metal, DNA-, iron/sulfur- and protein-binding.

econtact

econtact predicts inter-residue contacts for structure assembly and refinement.

Mandatory arguments:

-i input_ali, where input_ali is a file that contains threading alignments constructed by ethread.

-o output_file, where output_file will contain the predicted inter-residue contacts and the corresponding confidence estimates.

Optional arguments:

-p prob, where prob is a probability threshold used to identify confident contacts. The default value is 0.35, which was optimized during extensive benchmarks.

emodel

emodel constructs three-dimensional models using Modeller from alignments generated by ethread.

Mandatory arguments:

-t fasta_file, where fasta_file is a target sequence in FASTA format.

-s psipred_file, where psipred_file contains secondary structure predicted for you target sequence by PSIPRED.

-i input_ali, where input_ali is a file that contains threading alignments constructed by ethread.

-o output_file, where output_file will contain the coordinates of structural models constructed by Modeller for your target.

erank-modeller

erank-modeller ranks structural models generated by Modeller and provides confidence estimates.

Mandatory arguments:

-i input_file, where input_file contains concatenated models in PDB format generated by emodel.

-o output_file, where output_file will contain concatenated models, which are ranked and assigned a confidence.

erank-tasser

erank-tasser ranks structural models generated by TASSER-Lite and provides confidence estimates.

Mandatory arguments:

-i input_file, where input_file contains concatenated models in PDB format generated by TASSER-Lite.

-o output_file, where output_file will contain concatenated models, which are ranked and assigned a confidence.

-a input_ali, where input_ali is a file that contains threading alignments constructed by ethread.

-s psipred_file, where psipred_file contains secondary structure predicted for you target sequence by PSIPRED.

-c clust_file, where clust_file contains cluster data reported by TASSER-Lite.

Optional arguments:

-m minim, where minim is a structure minimization flag and should be either 0 (no mimization), 1 (remove atom clushes), 2 (refine loop regions) or 3 (refine loops and secondary structure regions). It uses minst from the Jackal package; the default value is 0. Setting it to 3 may result in more physical models.

eexctract

eextract extracts individual models from concatenated files generated by erank-modeller and erank-tasser.

Mandatory arguments:

-i input_file, where input_file contains concatenated models in PDB format.

-o output_file, where output_file will contain a single model in PDB format.

Optional arguments:

-n mod_num, where mod_num is the rank of a model to be extracted. The default value is 1, which corresponds to the top-ranked model.

-a align, where align is a flag that when set to 1 will center the extracted model on the origin of coordinate system and align its mass-weighted principal axes onto x, y, z. The default flag is 0.

econv-*

These scripts convert output files generated by individual threading programs to a format that is recognized by eThread, see the next section.

 

© Michal Brylinski
This website is hosted at the CCT