Note: This page is for LegacySEWING.

This is an outdated page. For information on the current version of sewing, please visit the Segment File Generation page. If performing hashed SEWING, visit the Structural Comparison of Substructures page. If you don't know what that means, start at the SEWING home page.


This application is used to generate model files and edge files to be used in the SEWING protocol. It is also used to convert SEWING edge files to a more memory-efficient binary format.


Science. 2016;352(6286):687-90 Design of structurally distinct proteins using strategies inspired by evolution. Jacobs TM, Williams B, Williams T, Xu X, Eletsky A, Federizon JF, Szyperski T, Kuhlman B.


The input files required by the sewing_hasher application depend on the mode being used to run the application (see the Command line options section below for details). Possible required inputs include a database (created using the FeatureReporters framework) to be used when generating SEWING models; a list of PDBs to be used when generating SEWING models; a SEWING model file; a SEWING edge file; and/or a PDB file.


The sewing_hasher's output again varies depending on the mode being used. Possible outputs include a SEWING model file, a SEWING edge file (text), a SEWING alignment file, or a binary SEWING edge file.

Command Line Options

SEWING-specific options

-sewing:mode   Mode used to run the application. Options are:
               generate: Used to generate a model file
               generate_five_ss_model: Generate a model file with five segments per model (under development)
               hash:     Used to generate an edge file
               pregenerate_alignments: Used to generate an alignment file
               convert:  Convert an edge file from text to binary format
               test:     Prints the name of the edge file and exits
-sewing:model_file_name Name of output model file (for generate mode) or input model file (for all other modes)
-sewing:score_file_name Name of SEWING edge file to output (hash mode) or input (align/convert modes)
-sewing:min_hash_score  Minimum number of overlapping atoms per segment to generate an edge
-sewing:max_clash_score Maximum number of clashing atoms (in the same bin but not superimposed) per alignment
-sewing:num_segments_to_match How many segments to overlay to generate alignments (default 1)
-sewing:match_segments  Which segments of each substructure should be scored by the hasher? 
                        (ex: 1 3 to score helices in a helix-loop-helix motif)
-sewing:max_models      Maximum number of models to hash (for testing only)
-sewing:starting_model  Specify a specific model ID to begin hashing
-sewing:num_procs       Number of processors to split up hashing with (hash mode only)

Structure Input Options

These options are required for certain sewing_hasher modes.

-in:file:l     In generate mode, can be used to provide a list of PDBs containing SEWING models.
               In align mode, can be used to list starting PDBs for which to generate alignment files
-inout:dbms:database_name  In generate mode, used to provide a Features database containing SEWING models
-in:file:s     In align mode, can be used to list a single starting PDB for which to generate an alignment file.


For instructions on using the sewing_hasher application to prepare input files for SEWING, see the Model generation and Model comparison with geometric hashing pages. Note: Only hash mode currently supports running on multiple processors

See Also