Skip to content

Latest commit

 

History

History
52 lines (39 loc) · 2.6 KB

README.md

File metadata and controls

52 lines (39 loc) · 2.6 KB

GlycoProteinBuilder

Uses GMML2 to add and adapt 3D structures of N-glycans and O-glycans onto glycoproteins. It can do this for Asn, Ser, Thr and Tyr.

General Concept

schematic

Notes

Project is under development, contact olivercgrant "at" gmail.com with queries. This code is intended to replace the glycoprotein builder currently available on glycam.org/gp. You can compile and run it locally. Has been tested on linux, but might install on both Mac and Windows.

Prerequisites

You'll need GMML2: Click here for installation instructions

Installation of GlycoProteinBuilder

The GlycoproteinBuilder will be compiled to gmml2/bin/gpBuilder after running the gmml2 make.sh script

Testing

Once compiled, a sample call to the program can look as follows

cd gmml2/tests
mkdir output
../bin/gpBuilder tests/inputs/017.GlycoproteinBuilderInput.txt output

You can run ../bin/gpBuilder --help for a list of options

usage: ../bin/gpBuilder [-h | --help] [-v | --version] [-p <value> | --num-threads <value>] 
                        input-file [output-directory]

Setup

Edit or create an input.txt file. See gmml2/tests/tests/inputs/017.GlycoproteinBuilderInput.txt for an example.

Required input:

A protein 3D structure
A Glycan 3D structure(s) or sequences in GLYCAM condensed nomenclature (just like the carb builder here: glycam.org/cb)
An input.txt, which contains:
    protein file name
    the protein residue numbers you want to attach to (no automatic detection of sequons)
    the glycan you want to attach in Glycam condensed sequence format.

Optional input : default values

  • NumberOfOutputStructures:1 -> each output structure is an independent sample, where shapes of glycans at each site are randomized according to their statistical likelihood.
  • persistCycles:5 -> how long the algorithm keeps trying to find an improvement in overlap. A higher value takes more time, but may better resolve overlaps.
  • seed: -> any number will yield a different result from other numbers, but will be reproducible across multiple runs. Delete this line for true randomness
  • freezeGlycositeResidueConformation:false -> if true, chi1 and chi2 angles of protein-glycan linkages will be preserved according to their shape in the protein file
  • deleteIncompatibleSites:false -> if true, algorithm will delete glycans where overlaps fail to resolve one at a time and start over, until all overlaps can be resolved. If false, all glycans will be preserved in each sample, even if the resulting structure has overlaps