NEWS & EVENTS

Accessing FcLigand features through a web browser:

The performance of calculations and graphical rendering was evaluated in a client-server web architecture using the Blazor framework (open-source framework from Microsoft to create web user interfaces based on components using C# and HTML). Blazor leverages WebAssembly technology developed by the World Wide Web Consortium since 2017. It delivers top-tier performance with a near-native instruction set for both x64 and ARM processors.

Based on this successful validation using unified C# code, we are now working on delivering full web-based graphical interfaces for FcBioisostere, FcLigand, and FcRebuilder software.

Promoting women’s rights with Cité des Droits des Femmes association:

In 2023 we were proud to initiate activities at Felix Concordia to promote human rights, which is one of the seven core subjects of the ISO 26000 international standard for social responsibility, a CSR guideline.

We became the first corporate sponsor of the Cité des Droits des Femmes project (https://CiteDesDroitsDesFemmes.fr), which aims to open the first museum in the Paris region dedicated to the history of women’s rights. Social rights in the workplace is one of their key themes.

A recent initiative was an interactive photo exhibition featuring slogans from International Women’s Rights Day, covering topics ranging from violence against women to gender equality in the workplace.

FcBioisostere new graphical features to better explore bioisosteric PDB based pairs:

This year we extanded the FcBioisostere software to better sort the precalculated 3D bioisosteric pairs extracted by our application from the PDB Protein Data Bank. Same topological pairs are now grouped together to help users assess a statistical suitability of the proposed fragment replacement onto their ligand of interest. Furthermore, Structural Chemical Features (Hbond donor/acceptor, charge, hydrophobicity, …) and protein of proposed bioisosteric pair can be displayed alongside the input ligand.

These and other user enhancements are now available in the latest FcBioisostere release.

Fig.: Replacement of hydrolysable methyl ester chemical moiety on ADAMs ligand query (pdb code 3SI9) with FcBioisostere ; on the left, upper view shows the input 2D or 3D query, lower table shows all query substructures to select the area to replace ; on the right, upper table shows all found replacement pairs, the selected pair is yellow was detected in 121 PDB pairs, lower view shows first the query ligand in white with the protein-fragment pairs in green and orange, then on the right, some of the 121 pairs of methyl-ester by 1,2-oxazole replacement.

Designing protocols with a workflow user interface:

In order to faster design new protocols for users in our C2P Chemo-Proteomic Platform, we investigated technical solutions to provide a sketcher to users to design predictive protocols similar to a workflow.

FcPipeline our new module under development provides a full set of input/calculation/output nodes that can be chained together to define any chemoproteomic protocols. The user interface dynamically displays the number of processed molecules at each stage of the workflow. The processing of ligand and protein flows is managed using iterator technology, which allows for immediate results for a given protocol, enabling users to verify the protocol’s validity without having to wait for the entire calculation to finish.

Any of our C2P API methods can be integrated as a separate node, with parallel processing enabled for that node if needed to improve performance. We plan to include all our chemo-proteomics C# methods in a future release of FcPipeline.

Fig.: FcPipeline protocol for fragmenting and 3D-hybridizing two sets of pre-superimposed ligands; red nodes represent input readers; orange and green nodes represent cheminformatics methods; gray nodes are viewers (all displayed on the left side of the view, with the 3D hybrid structures shown below, using red or green substructures to indicate the source ligand)

Proof Of Concepts to annotate protein surfaces and generate drug candidates:

Since the beginning of the COVID-19 pandemic crisis, the scientific community has generated an amazing amount of data especially in structural biology with more proteins together with cocrystallised ligands of fragment probes. We ran our 3D intermolecular surface miner protocol in to first superpose the PDB materials having similar molecular surface, then combined in FcLigand the superpose ligand/fragment materials into protein pockets, and finally annotated the generated molecules by matching to PDB/Pubchem/ChEMBL repositories.

On Spike protein that project from SARS-CoV-2 surface (pdb code 7JV4), we compared the full 3D surface to retrieve PDB proteins showing similar local network of 3D interactions. The resulting superposition highlights epitopes positions on the RBD Rigid Body Domains targeted by actual vaccines.

Fig. Epitope detection on SARS-CoV-2 Spike protein: spreadsheet on top summarizes 19 non-SARS-CoV-2 proteins that interact with Spike protein ; 3D viewer displays in white the 7JV4 query backbone and 3 hits in ball rendering, n°494 (7LM9 in blue), n°604 (2DD8 in orange), and n°848 (7JN5 in green vert) ; only antibody part of hits are displayed, network of 3D shared structural chemical features (Hbond donor/acceptor, charge, hydrophobicity, aromatic, …) for the 3 hits highlights putative epitope surfaces.  

Onto NSP 5 Main Protease (pdb code 6Y2F), we collected 209 PDB ligands having 3D similar interactions with the query and no strong steric clashes, then 3D-combined by detecting the Maximal Common Substructure to obtain 2320 hybrids ; 24 were 2D-matching to PDB ligands, 92 to Pubchem molecules, 2 to ChEMBL molecules.

Fig. FcLigand 3D-hybrisdisation of ligand structures from 209 PDB complexes where 3D similar interaction surfaces were detected and superposed onto the 6Y2F binding site ; the displayed hybrid corresponds to CHEMBL 213054 by 3D hybridisation of 3SZN, 3AVZ and 7C8T co-crystallised ligands (shared detected scaffold is in gray)

Additionally, we used FcRebuilder on MPro to dock 212 ChEMBL active molecules on SARS-CoV-2 targets ; our software first retrieved 900 protein-ligand structures with a 3D interaction surface similar to our selected 6Y2F target.  These superposed ligands were then used to 3D-rebuild the ChEMBL subset into the binding site; the predicted poses positioned at least 90% of the structure.

Fig. Rebuilding/Docking ChEMBL212080 into 6Y2F MPro with FcRebuilder software ; left view shows after 3D superposition into binding site five cocrystallised ligands from 7JW8 in green; 7M00 in  magenta, 2A5K in yellow, 7LBN in brown, and 6LKA in blue ; middle view shows the fragment contribution to rebuild ChEMBL 212080 molecule ; right view displays the pause in the MPro target.

These 3 POCs highlight the benefit of our PDB based drug design protocols with our software technology

Code standardization in C# for agile developments:

To streamline our software development processes, we decided this year to migrate some of our components to C#, our primary programming language. By bypassing dynamic serialization, we gain advantages in speed, memory usage, and enable full polymorphism for data types throughout the code.

Twenty years after it first release, C# is a now a powerful high-level programming language supported on different plateforms. At Felix Concordia, we appreciate the integration with Visual Studio IDE for faster code development, the strong typing to prevent bugs, and parallelization methods to provide high performance on multi-core processors.

We are pleased to announce the release of FcLigand 2.01, FcBioisostere 2.01, FcRebuilder 2.01, and FcTools 2.01, all in full 64-bit native architecture.

Poster presentation at the 2017 Groupe de Graphisme & Modélisation Moléculaire conference (Reims, France): 

Fast software development in drug design with the C2P-API toolkit : Rebuilding ligand bound conformation from PDB subpocket superpositions

Abstract: Literature in the FBDD field reported cases where the binding location and orientation of small co-crystallized fragments is preserved when these fragments are chemically evolved into larger ligands. Crystallography also revealed the central role of some weak fragment binders. With the benefit of the PDB growth, pocket superposition illustrated some ligand substructures having conserved binding modes towards multiple targets. Recently, we showed that 48% of 2,241 ligands structures uniquely represented in the PDB could be reproduced at 80 % of their atom coordinates within a 1.0 Å deviation by using chemical material from 3D pocket superpositions.

The development of an heuristic to rebuild ligand bound conformation from PDB pocket superpositions requires at least methods to detect Maximum Common Substructures between the compound to 3D-rebuild and all PDB ligands from superposed pockets, and methods to 3D-hybridise those MCS fragments. It was possible to develop a rebuilder prototype including a graphical interface in about 3 months by combining (a) the richness of the C2P-API Chemo-Proteomic Advanced Programming Interface, (b) the agility of the .Net C# language including reconfigurable dictionaries to include molecular rules, quasi-automatic parallelization and safe memory management, and (c) the comfort of the Microsoft Visual Studio Community IDE in developing graphical interface and debugging software.

Our auto-converging knowledge-based iterative hybridiser software is 3D-rebuilding as much as possible any input 2D-ligands into any biostructural binding site. We selected inhibitors co-crystallized in multiple PDB proteins to validate our protocol.

Click here to download the PDF poster

Pre-release of the Felix Concordia cheminformatics Advanced Programming Interface:

Today July 1st 2013, Felix Concordia is providing access to the SupBiotech engineer school (http://www.supbiotech.fr/) on an Advanced Programming Interface dedicated to cheminformatics. This set of documented informatics classes represents a consistent architecture to develop on demand new predictive methods/procedures on ligands and/or biological targets. Written in C#, this API takes advantage of the richness Microsoft .Net platform and of the agile Microsoft Visual Studio Community IDE in developing graphical interface and debugging software (Express and Community versions are accessible for free for students). Our Felix Concordia API is provided with a Doxygen-based documentation to search by keywords the required function/method/class/attribute. In such environment, Supbiotech’s students will be able to write in few days training projects in cheminformatics.

The C2P Chemo-Proteomic Platform initiative to address new challenges in drug design:

On July 1st 2013, MEDIT and Felix Concordia have launched the C2P Chemo-Protein Platform initiative to provide a complete software solution to better mine altogether biostructures, structure activities and chemical libraries at PDB, Pubchem and multi-million compound levels. We strongly believe this is opening new perspectives in drug discovery and target profiling by combining knowledges from experimental data with existing predictive models.

C2P architecture allows to cross-mine with interactive 1D/2D/3D/ND queries a data repositories of biostructures, structure activities and chemical libraries. Navigation conjointly in binding sites, ligands, subpockets, fragments 2D/3D similarities including anti-targets, unwanted chemistry, and in-house data provides a new smart experience to highlight selectivity mechanisms. C2P initiative now includes MED-SuMo-GUI (Medit SA) to detect/superpose similar 3D protein surface interactions, MEDP-Fragmentor (from MEDIT) to deconvolute protein-ligand structure in pocket-fragment interactions, MEDP-Site-Classifier (Medit SA) to classify the full Protein Data Bank according to pocket similarities, FcLigand (Felix Concordia SARL) to explore 1D/2D/3D ligand similarities, FcBioisostere (Felix Concordia SARL) to better profile your compound by using  bioisosteric replacement, and the C2P-API open source advanced programming interface.

MEDIT SA, a privately held software company, has developed a large experience on solutions to superpose protein binding sites for functional annotation, off-target identification, drug repurposing, and fragment-based drug design. Felix Concordia SARL is providing consistent software components to mine large set of data in life science.

First sale of FcBioisostere software:

On November 11th 2011, the Center for Biophysical Sciences and Engineering of the University of Alabama acquired multiple software licenses to use FcBioiostere on the whole campus.

FC-Bioisostere  opens access to 3D bioisosteric replacement onto a lead molecule to find chemical groups having similar 3D biological interactions. While maintaining target potencies, it helps Chemists to optimize additional properties in pharmacokinetics and metabolic response, and/or to escape from existing patents by selecting alternative equivalent chemical groups.