Release Notes

Version 2024.1

This version adds the following:

  • SILCS-MC using GPU (Nvidia CUDA), performance increased by 700%

  • SILCS-MC: added slab mode for docking along user-defined path

  • New SilcsBio GUI additions

    • SILCS-RNA in GUI

    • SILCS-Membrane Protein in GUI

    • SILCS-MC Pose Evaluation in GUI

    • CGenFF Suite in GUI with basic functionality

  • New CLI additions

    • SILCS-MC Pose Evaluation in CLI

    • Bayesian Machine Learning module (SILCS-BML) in CLI

    • SILCS trajectory scanning module for finding the protein conformations that best complement docked ligand poses in CLI

    • Pharmacophore screening (Pharmer) in CLI

    • CGenFF-covalent module in CLI

    • MD module in CLI

      • Standard MD: setup, run and analyze

      • MD with covalent ligand

      • MD with single probe

      • MD for membrane protein

as well as bug fixes and stability improvements.

Version 2023.1

This version adds the following:

  • GCMC sampling utilizing GPU resources. Prior versions utilized GPU resources for the MD portion of the GCMC/MD sampling in SILCS simulations whereas GCMC sampling was done using CPUs only. The new GPU GCMC code can make runs 50% faster overall while also enabling more efficient probe sampling due to algorithmic improvements for mu_ex oscillation.

  • SILCS protocol specifically tailored for RNA targets

  • SILCS-Biologics workflow setup, run control, and data analysis from the SilcsBio Graphical User Interface (GUI)

  • SILCS support for membrane proteins other than GPCRs, such as transporters and ion channels

  • SILCS-Hotspots output validation capability by docking of a diverse subset of FDA-approved drugs

  • Mechanism for checking internet connection, server connection, and SILCS server version compatibility from within the SilcsBio GUI

  • Integrated SILCS-Biologics documentation access from within the SilcsBio GUI

  • Updated graphical user interface design

  • Compatibility with GROMACS version 2022.x

  • Improved jobs control within the SilcsBio GUI

  • Scripting to subtract one set of FragMaps from another, including automatic handling of differences in FragMap grid dimensions. This can be useful for qualitative analysis of functional group binding preferences between two homologous proteins.

as well as bug fixes and stability improvements.

Version 2022.2

This version adds the following:

  • SilcsBio Graphical User Interface (GUI) support for Halogen SILCS simulation

  • SilcsBio GUI support for inclusion of Halogen SILCS FragMaps in ligand posing and scoring for SILCS-MC: Docking and Pose Refinement and SILCS-MC: Ligand Optimization

  • Support for custom temperature choice for SILCS simulations

  • Support for using existing setup/trajectory files for starting SILCS simulations via the SilcsBio GUI

  • Expanded SilcsBio GUI visualization capabilities including: atomic GFE values and their sums for ligands scored with SILCS-MC; comprehensive atom selection options for visualization; and residue labels on proteins

  • Usability improvement to the SilcsBio GUI including: showing project creation dates; ability to sort by date or name during file/directory selection; and progress bar for file transfers

  • CGenFF program is version 2.5.2 and CGenFF parameters are version 4.6

as well as bug fixes and stability improvements.

Version 2022.1

This version adds the following:

  • SilcsBio Graphical User Interface (GUI) support for SILCS-Hotspots

  • Distance measurement between SILCS-Pharm pharmacophore features in the SilcsBio GUI

  • Charting results from SILCS-MC jobs (ligand optimization, docking, pose refinement, hotspots) prior to full job completion, and recalculation of results at any time through the SilcsBio GUI

  • Selection of number of processors and number of independent runs for SILCS-MC through the SilcsBio GUI

  • SILCS-Hotspots report generation script update to show additional measurements

  • SILCS-MC docking and pose refinement expansion to include Halogen FragMaps

  • SILCS simulation with iron ions can be with +2 (default) or +3 charge

  • Dimethylether replaces acetaldehyde in Standard SILCS simulations

  • CGenFF program update to version 2.5.1 and CGenFF parameters update to version 4.6

    The CGenFF program version 2.5.1 can optionally retain partial charges from the input Mol2 file instead of assigning them. CGenFF parameters version 4.6 is updated with training that adds numerous new compounds:

    Model compounds for nonstandard amino acids (206 new residues encompassing 87 new bond parameters, 437 new angle parameters, 1698 new dihedral terms, and 34 new improper dihedral parameters) [19]

    Flavin and flavin-related model compounds

as well as bug fixes and stability improvements.

Version 2021.1

This version adds the following:

  • The SILCS-Biologics method for excipient screening for biomacromolecular therapeutics

  • CGenFF program update to version 2.5 and CGenFF parameters update to version 4.5

    CGenFF coverage has been extended to the following compounds that have been explicitly parameterized and validated, allowing CGenFF to generate parameters for more diverse compounds.

    ASBB: (1-[(2-aminoethyl)sulfanyl]butan-1-one): charges and bonded parameters

    2-phenylthiazole and 5-methyl-3-phenyl-1,2,4-oxadiazole: bonded parameters

    Fentanyl: bonded parameters [17]

    (E)-1,2-di-p-tolyldiazene: bonded parameters [18]

as well as bug fixes and stability improvements.

Version 2020.2

This version adds the following:

  • SilcsBio Graphical User Interface (GUI) improvements for file and directory selection allowing for input files to be chosen from remote servers as well as the local computer

  • SilcsBio Graphical User Interface (GUI) support for visualization of Halogen SILCS FragMaps

  • A new plug-in for visualizing SILCS FragMaps in MOE

  • CGenFF program and parameters update to version 2.4.0

    CGenFF coverage has been extended to amide bases and molecules containing boron. The functional groups in these molecules were not previously accessible in CGenFF. The following compounds have been explicitly parameterized and validated, allowing CGenFF to generate parameters for more diverse compounds.

    N-methyl acetamide (deprotonated amide); N-ethyl acetamide (deprotonated amide); N-methyl benzamide (deprotonated amide); N-phenyl acetamide (deprotonated amide)

    Methyl boronic acid (neutral, -1, -2); Ethyl boronic acid (neutral, -1, -2); Phenyl boronic acid (neutral, -1, -2)

    Additionally, 112 naturally-occurring modified nucleotides, especially modified bases with heterocycles not previously covered by CGenFF, have been parametrized. Quantum mechanical calculations on model compounds, including geometries, dipole moments, and interactions with water, provided target data. Selected parameters were validated with extensive molecular dynamics simulations. Details of the parameter optimization and the complete list of nucleotides can be found in [16].

as well as bug fixes and stability improvements.

Validation of the SILCS-HotSpots approach has been published [13].

Version 2020.1

This version adds the following:

  • The SILCS-Hotspots method for identifying all potential ligand binding sites on a protein

  • SilcsBio Graphical User Interface (GUI) support for SILCS-Pharm

  • SilcsBio Graphical User Interface (GUI) support for SILCS-MC Docking and Pose Generation

  • Performance improvements to GCMC-MD

  • CGenFF program and parameters update to version 2.3.0

    CGenFF version 2.3.0 adds explicit parametrization for the following molecules. In prior versions, functional groups in these molecules were accessible through analogy to related functional groups. Now, explicit parametrization and validation yields more accurate treatment and decreases the associated CGenFF penalty scores.

    1H-tetrazole; 5-methyl-1H-tetrazole; 5-ethyl-1H-tetrazole

    2-oxetanone; 3-oxetanone

    ammonium; dimethylammonium; trimethylammonium (Note: Protonated amine parameters were previously based on methylammonium. While the present explicit parametrization of secondary and tertiary amines leads to smaller penalty scores and improvements in performance, electrostatic interactions will continue to be dominated by the +1 monopole.)

    1-butyne; 1-pentyne; 1-hexyne; 1-heptyne; 1-octyne; but-1-ene-3-yne (Note: Alkyne parmeters were previously based on ethene and propene. Extension to longer alkynes and the ene/yne combination validates the parameters and leads to improved treatment of the intramolecular interactions.)

    CGenFF version 2.3.0 also includes improved halogen–protein interactions. Quantum mechanical calculations on chloro- and bromobenzene with model compounds representative of protein functional groups were used as target data to optimize atom-pair specific Lennard-Jones parameters for selected atoms in the model compounds. Application of the parameters in molecular dynamics simulations of eight ligand-protein systems demonstrated systematic improvement in interaction geometries.

as well as bug fixes and stability improvements.

Version 2019.2

This version adds the following:

  • The SILCS-Pharm method for generating 3D receptor-based pharamcophore models from FragMaps in an automated manner

as well as bug fixes and stability improvements.

Version 2019.1

This version adds the following:

  • The SILCS-Pharm method for generating 3D receptor-based pharamcophore models from FragMaps in an automated manner (early access)

as well as bug fixes and stability improvements.

Version 2018.2

This version adds the following:

  • A new Graphical User Interface (GUI) capable of preparing and launching SILCS, SILCS-MC, and SSFEP jobs on remote computing clusters and analyzing and visualizing the job outputs

  • A new CGenFF Parameter Optimizer with functionality for dihedral parameter fitting

as well as bug fixes and stability improvements.

Version 2018.1

This version adds the following:

  • A new Graphical User Interface (GUI) capable of preparing and launching SILCS, SILCS-MC, and SSFEP jobs on remote computing clusters and analyzing and visualizing the job outputs (early access)

  • Improved GPCR and other transmembrane protein support for SILCS

  • Improved GPCR and other transmembrane protein support for SSFEP

  • CGenFF program and parameters update to version 2.2.0

    CGenFF version 2.2.0 extends support to a large variety of drug-like molecules to be used routinely in computer-aided drug design projects. Specifically, 1) it improves treatment of halogen bonds by introducing lone-pairs onto the halogen atoms of aromatic systems. 2) Support is extended to four-membered oxetane and glycoluril. 3) Improved predictions with partial charge distributions around ammonium ions and primary amines.

  • A new CGenFF Parameter Optimizer with functionality for dihedral parameter fitting (early access)

as well as bug fixes and stability improvements.

Version 2017.2

This version adds the following:

  • GPCR support for SILCS

  • GPCR support for SSFEP

  • CGenFF program and parameter update to version 2.1.0

    CGenFF version 2.1.0 extends support to S-P bond found in the GTP-gamma like molecules. Bonded parameters and charge-distribution along the S-P bond were modeled using the CHARMM nucleic acid force-field patches for mono- and di-thio substitutions (toppar_all36_na_reactive_rna.str). Three new atom-types have been added to the CGenFF force-field : SG2P1, SG2P2 to support the mono- and di-thio substitutions, along with OG2S1 to model the terminal oxygen connected to the S-P bond.

as well as bug fixes and stability improvements.

Version 2017.1

This version is the initial release. This version includes the following packages:

  • SILCS command line interface

  • SILCS-MC command line interface

  • SSFEP command line interface

  • CGenFF program and parameter version 2.0.0

    CGenFF version 2.0.0 is the second release of CGenFF, extending support to a larger variety of drug-like molecules to be used routinely in computer-aided drug design projects. Specifically, it improves treatment of halogen bonds, by introducing lone-pairs onto the halogen atoms of aromatic systems. Additionally, support is now extended to four-membered oxetane and glycoluril moieties. This is driven largely by improvements in the newly released version 4.0 of CGenFF force field.