Subhas
Chakravorty US Citizen
https://www.linkedin.com/in/subhaschakravorty
sjchakravorty@hotmail.com
215-900-0116
Primary Work is therapeutic projects support: Data dashboards, molecular design & decision, support of 4 Kinase projects supporting individual project managers for medicinal chemistry design decisions to both bench medicinal & screening scientists and the executive leadership.
Expertise in a variety of algorithms, methods, and unsupervised learning techniques, with a focus on their application to chemical & biological data (HTS and OMICS)
Experienced both as a consultant and in house FTE scientist and leader.
Hands on experience in setting and running AI and Predict first ecosystem.
Hands on design experience of HTS collection, hit marking, hit to lead, and lead optimization.
Hands on accurate building block identification, reaction enumeration and scoring.
Reaction based and R-group based enumeration (profiling and big data reduction via similarity of R-group, domain applicability ΰ AI/ML SAR: idea generation & project support. Molecular Matched Pair visualization via spotfire.
Use of MOE software for automated docking and molecular modeling (and Maestro)
Hands on ETL (extract, transform and load data) activities into data lake or database of chemical structure data and biological data. Registrar duty: Compound registration into Dotmatics registration system.
Use of Schrodinger Maestro and Jaguar for physics based conformational modeling, ab initio calculations.
Hands on machine learning expertise: Solid experience with AI and machine learning tools libraries and frameworks Python, Pipeline Pilot, Perl and C both building tools and applications.
Statistical design of experiments (DoE), multiparameter optimization, AI/ML SAR for lead optimization; formal background and extensive work experience in medicinal chemistry support, informatics & AI/ML computations & methodology development, computer science, ab initio electronic structure (Jaguar, Gaussian, Psi4) and numerical intensive computations. Density Functional Theory and Molecular calculations expert.
Numerical algorithms and deployment scientific computer programs for Spotfire, Excel and Web visualization using Pipeline Pilot, and Python for each therapeutic project.
Ab Initio Quantum Chemistry and DFT (Ph.D., Postdoctoral, Industrial Code development and applications experience)
Advanced computer skills including programming experience with python, perl and C++,
both in UNIX & Windows environments. Budgeting, maintenance, installation, and monitoring use of Comp Chem software. Rapidly developing and writing bespoke computer code as needed for business.
Principal (working with Client involved in RNA therapeutics discovery research)
Currently consulting with Skyhawk Therapeutics, design of screening collection, hit to lead and lead 3D modeling work.
Use of Maestro software for conformational, molecular modeling and ab initio calculations for medchem clients
HTS Screening Deck Design: Wrote computer code in python (rdkit, sklearn, rapids) clustering and t-sne calculations.
Principal Research Investigator
Spotfire drug discovery project dashboards (Live Design-like view)
Daily updates of Oracle tables for registered & virtual compound properties and AI models, structure alerts, compound synonyms and target Information.
Compound registration (Dotmatics)
Medchem SAR support: R-group deconvolution project view, top compounds project-based card view spotfire reports
Built, deployed, and auto improve ~45 AI/ML SAR models for Biochemical, Cellular, stability, PK ADME & TOX assays which auto populate Dotmatics compound views.
Worked on the lead to candidate optimization of (targets and candidates that have now been recently disclosed by Kinnate Biopharma)
o KIN-8741, c-MET (AI/ML SAR R-group optimization and Array, Dashboard profile)
o CDK4 selective program (and other CDK) (AI/ML SAR R-group optimization and Array, Project dashboard profile, alignment of CDK family and autodocking and autoscoring of idea ligands onto Xray structures.)
o KIN-7136, a brain-penetrant MEK inhibitor for MAPK pathway-driven solid tumors (Project dashboard profile)
o Selective and a brain-penetrant RAF inhibitor (2 compounds) (AI/ML SAR R-group optimization and Array. Project dashboard profile)
New Targets research (Drug and OMICS data mining and analysis)
Maintenance on AWS of computational chemistry software, installation, and licenses (ACD, MOE, Pipeline Pilot, Spotfire, Chemdraw)
Auto dock with MOE all registered leads/ideas to several related receptors into a new MOE database. Combine several MOE dbs. into a single and extract docked scores.
Investigator, Computational Chemistry Dept (2019-2021)
Computational Biology Data Manager, Target Sciences, Computational Biology Dept
(2016-2019)
Investigator, Computational Chemistry Dept (2006-2015)
Developed & Used a One Prot Reaction Enumerator: AI/ML SAR & Multi Parameter Optimization; Spotfire visualization & Building Block Requisition Spreadsheet.
GSK building-block analysis, computer chemical reaction enumeration and multi-objective ranking of building blocks and products. Maintain Building Block selection tool for Med chemists from various Data sources including internal Discovery Supply, Frontier and Chemstores. Building Block Replenishment and Acquisitions expert for GSK.
Bioinformatics Projects :Target focused compound Identification (external data sources of drugs and GSK biological assay data). Created detailed targets ↔ compounds annotation tables: TaCos (widely used in GSK). Continuously update this primary data for GSK use. An example of use of the TaCos compilation was to recently identify suitable drug compounds for testing in COVID-19 assays. Developed computer programs that integrate multiple vendors and public repositories.
Cheminformatics HTS screening progression support for 6-7 therapeutic programs per year. Also, developed and used novel hit marking tool in HTS.
Identified, assembled the suitable compounds in the GSK HTS collection in response to the request from sample management. Implemented screening, chemistry and computational chemistry business rules and sample replenishment ideas into the workflow.
Expert Pipeline Pilot programmer of cheminformatics, screening, sample informatics and proteomics workflows and the pipeline pilot application administrator.
Expert use of Discovery Studio modules for conformational modeling specifically 3D pharmacophores
Scientific information curation and crafting computational database queries, complex joins, data mining, data analysis and Spotfire reports.
Built project specific and generally applicable AI/ML SAR models and applied them to virtually enumerated libraries for many therapeutic programs.
Building Block and substructure searches suitable for specific reactions and sourcing for GSK medicinal and DNA library chemistry community. enumeration of chemical reactions, virtual libraries, and screening for DNA encoded libraries.
Focused screening compound sets: Index set, natural product, sugar, kinase, marketed drug, biological lead like and experimental drugs.
Fold change analysis, proteomics, and microarray: implemented several proteomics and bioinformatics workflows on pipeline pilot, proteomics, microarray gene expression.
Co-authored 2 patents using AI/ML SAR to build a predictive model for RIP2 therapeutic drug discovery program.
Shaping and data ingestion of terabyte sized Hadoop based impala tables and making useful and on-demand Spotfire dashboard visualizations. These are public datasets like Blueprint, GTEX, Human Disease, TCGA and others. This and the underlying methodology of on-demand data population is now widely used in GSK.
Curating and maintenance of in-house structure correspondence tables of compounds between various public databases
QSAR AI/ML Software Package: AI/ML software to deliver >thousands of models for target with automatic selection of top models using multi objective optimization. Also, I presented this work in Biovia User Group meeting and internally in GSK. (https://3dsbiovia.com/events/hosted-events/ugm/2019/agenda.html)
Established the connection between OMICS data and traditional SAR AI/ML and applied this methodology for the repurposing of marketed drugs.
Affinity selection mass spectrometry (ASMS) and
GSK HTS hit analysis and comparisons for several targets, validating that ASMS
is a useful primary screening strategy.
Demonstrated the ability to enumerate,
characterize, profile huge ELT libraries from sequence «
building block tables.
Worked on Amgens HLEE project (Holistic Lab execution environment): Developed several specific parser software for instrument data acquisition for ELN systems with Pipeline Pilot. (e.g., Guava, XRPD, CARY 50, Forte Bio, BGA, Mettler Balances, AKTA, TGA instruments). This required collaborative code and protocol development with Biovia developers.
Performed clients instrument inventory analysis for ELN from Maximo (IBM database product containing Amgens corporate inventory)
Performed Integration of Workforce Data into ELN from the client corporate database
Technical presentations and customer 1-1 meetings and development of whitepapers.
Pipeline Pilot, Material Studio, Discovery Studio installation and maintenance on Microsoft Azure Cloud (Linux and Windows instances)
Provided applications, sales and technical support for customer accounts and business relationships with investors. Responsible for providing clarifying answers to investment analysts.
Responsible for preparation of SOW, RFI/RFP, Term Sheet documents
Designed, developed, authored several large sample size clustering packages for molecular data and their applications: Relocation (K-means), Jarvis Patrick, and Hierarchical clustering in C.
Algorithm and program development of Machine Learning methods (KNN, CQSAR) Column Diversity package and molecular descriptor generation and management C packages.
MIX (Merck Molecular Modeling Package for UNIX) Release manager and maintenance person. Conduct timely releases, developer meetings i.e., the MIX team, using the CVS code development system.
Excel Macro for automated graphs and spreadsheet for a reaction library design for outsourcing
Conformational modeling and program support for Men. B therapeutic project using Macromodel (Maestro), Cerius2 and Open Eye, Gaussian software.
Wrote several new C modules in the Cerius2 product (Drug Discovery Work Bench: coverage diversity, column diversity, fast descriptors, SMILES, SMARTS representation and use.
Debugged several older modules (mol representation of SMILES, SMARTS) in C and provided customer support.
Developed, and rewrote MS ACCESS, Excel, and Visual C program: PC-FORMULATE 3.1 (Windows 3.1/95) an AI/ML program for paint formulations.
Developed and authored Polymer Calculator, Particle Size Calculator, Enzyme-Ligand Binding IC50 kinetics calculator (Visual C and Basic applications)
Support and Programmatic support for SIMULSOV (Dow product) for Reaction Kinetics modeling (IBM VM, AIX)
Development of torsional Amber force field parameters for diacyl hydrazides. (Maestro/Macromodel, Batchmin, Gaussian, Jaguar)
Postdoctoral Research 1990-1996, Quantum Chemistry and Density Functional Theory (DFT) Indiana University, Bloomington & Computer Science coursework (IU and Purdue: 2 semesters undergraduate and graduate level)
Postdoctoral Research Quantum Chemistry and Density Functional Theory (DFT) 1987-1990, IBM Kingston, Kingston. NY
Ph.D., Quantum Chemistry and Density Functional Theory (DFT) 1982-1987, University of Poona, & Computer Science coursework (3 semesters undergraduate and graduate level)
Extant Informatics tools: Spotfire, R, Pipeline Pilot, Python
Data Science: numerical analyses, multi-objective optimization and modern machine learning methods, algorithm design and data structures.
Use of modules in Python: pandas, numpy, scipy, matplotlib, sympy, statsmodel, scikit learn
Also, developed in Perl, C, Pipeline Pilot-Pilotscript, FORTRAN, Visual Basic, Pascal, VB script,
Formal background: Computer Science,
Algorithms, Data Structures, Numerical Analysis (aka AI/ML)
Computational Chemistry software:
MELD, Gaussian, HONDO, Pipeline Pilot, Discovery Studio, Schrodinger, Open Eye,
MOE
DATABASES: Direct experience with Oracle SQL, MySQL, Hadoop SQL, SQLite
OPERATING SYSTEMS: Linux, UNIX (and VM, MVS, VMS,)
Formal industrial & academic code development both in Ab initio, DFT and Molecular Force fields and Cheminformatics.
1. Patent INHIBITORS OF RAF KINASES, filed (2023), Toufike Kanouni Jason M. Cox John Tyhonas Robert S. Kania Subhas J. Chakravorty Young K. Chen
2. Life and Science of Clemens C. J. Roothaan SR Gadre and S Chakravorty, Resonance 26, 737-755 2021
3. Nuisance compounds, PAINS filters, and dark chemical matter in the GSK HTS collection Subhas J Chakravorty, James Chan, Marie Nicole Greenwood, Ioana Popa-Burke, Katja S Remlinger, Stephen D Pickett, Darren VS Green, Martin C Fillmore, Tony W Dean, Juan I Luengo, Ricardo Macarrσn SLAS DISCOVERY: Advancing Life Sciences R&D 23 (6), 532-545
4. Development of a High-Throughput Cul3-Keap1 Time-Resolved Fluorescence Resonance Energy Transfer (TR-FRET) Assay for Identifying Nrf2 Activators DD Poore, G Hofmann, LA Wolfe III, H Qi, M Jiang, M Fischer, Z Wu, SLAS DISCOVERY: Advancing Life Sciences R&D 24 (2), 175-189
5. Selective class IIa histone deacetylase inhibition via a nonchelating zinc-binding group. Mercedes Lobera, Kevin P Madauss, Denise T Pohlhaus, Quentin G Wright, Mark Trocha, Darby R Schmidt, Erkan Baloglu, Ryan P Trump, Martha S Head, Glenn A Hofmann, Monique Murray-Thompson, Benjamin Schwartz, Subhas Chakravorty, Zining Wu, Palwinder K Mander, Laurens Kruidenier, Robert A Reid, William Burkhart, Brandon J Turunen, James X Rong, Craig Wagner, Mary B Moyer, Carrow Wells, Xuan Hong, John T Moore, Jon D Williams, Dulce Soler, Shomir Ghosh, Michael A Nolan, Nat Chem Biol 2013 May 24;9(5):319-25.
6. CCR2 receptor antagonists: Optimization of biaryl sulfonamides to increase activity in whole blood Gren Z. Wang, Pamela A. Haile, Tom Daniel, Benjamin Belot, Andrew Q. Viet, Krista B. Goodman, Deyou Sha, Sarah E. Dowdell, Norbert Varga, Xuan Hong, Subhas Chakravorty, Christine Webb, Carla Cornejo, Alan Olzinski, Roberta Bernard, Christopher Evans, Amanda Emmons, Jacques Briand, Chun-Wa Chung, Ruben Quek, Dennis Lee, Peter J. Gough, Clark A. Sehon, Bioorganic & Medicinal Chemistry Letters 21 (2011) 7291-7294
7. Patent: WO 2011120026. PYRAZOLYL-PYRIMIDINES AS KINASE INHIBITORS Casillas, L., Chakravorty, S., Eidam, P., Haile, P., Hughes, T., Lakdawala, A., Leister, L., Miller, N., Rahman, A., Sehon, C., Wang, G., Zhang, D. PCT Int. Appl. (2011)
8. Patent: WO 20111120025. INDAZOLYL-PYRIMIDINES AS KINASE INHIBITORS. Casillas, L.N., Chakravorty, S.J., Charnley, A.K., Eidam, P., Haile, P.A., Hughes, T.V., Jeong, J.U., Jianxing, K., Lakdawala Shah, A., Leister, L.K., Marquis, R.W., Miller, N.A., Price, Daniel J., Sehon, C.L., Wang, G.Z., Zhang, D.
9. Web enabling technology for the design, enumeration, optimization and tracking of compound libraries Bradley P. Feuston, Subhas J. Chakravorty, John F. Conway,J. Christopher Culberson, Joseph Forbes, Bryan Kraker, Patricia A. Lennon, Craig Lindsley, Georgia B. McGaughey, Ralph Mosley, Robert Sheridan, Mario Valenciano and Simon K. Kearsley, Curr Top Med Chem. 2005;5(8):773-783.
10. Reagent Selector: Using Synthon Analysis to Visualize Reagent Properties and Assist in Combinatorial Library Design, Ralph T. Mosley, J. Christopher Culberson, Bryan Kraker, Bradley P. Feuston, Robert P. Sheridan, John F. Conway, Joseph K. Forbes, Subhas J. Chakravorty, Simon K. Kearsley. J. Chem. Inf. Model. 2005, 45, 5, 14391446
11. Diversity and Coverage of Structural Sub libraries Selected Using the SAGE and SCA Algorithms, Charles H. Reynolds, Alexander Tropsha, Lori B. Pfahler, Ross Druker, Subhas Chakravorty, G. Ethiraj, and Weifan Zheng, Journal of Chemical Information and Computer Sciences; 2001; 41(6); 1470-1477.
12. Improved AMBER* torsional parameters for the N-N rotational barrier in diacylhydrazines, Subhas Chakravorty and Charles H. Reynolds, J. Mol. Graphics and Modeling, 17, 315, 1999.
13. Coupled cluster calculations of equilibrium geometries, harmonic vibrational frequencies and the barrier height of ethane, Roberto-Neto-O; Chakravorty-S; Machado-FBC, J Mol Structure Theochem. (2002); 586: 29-34.,
14. Ab Initio Calculations on XFn ( X = I -, Xe, Cs +, Ba ++ ) and n = 1,2,4,6 molecules, FBC Machado, TK Ghanty, S Chakravorty and ER Davidson, Int. J. Quant. Chemistry 81, 238 (2001).
15. Electron Momentum Spectroscopy Experiments and Calculations for the Production of the Excited States of He+ and H2+, N. Lermer, B.R. Todd, N.M. Cann, C.E. Brion, Y. Zheng, S. Chakravorty and E.R. Davidson, Can. J. Phys. 74, 748-756 (1996).
16. Orbital Momentum Profiles and Binding Energy Spectra for the Complete Valence Shell of Molecular Fluorine, Y. Zheng, C.E. Brion, M.J. Brunger, K. Zhao, A.M. Grisogono, S. Braidwood, E. Weigold, S.J. Chakravorty, E.R. Davidson, A. Sgamellotti and W. von Niessen, Chem. Phys. 212, 269-300 (1996).
17. Imaging of the HOMO Electron Density in Cr (CO)6, Mo(CO)6 and W(CO)6 by Electron Momentum Spectroscopy: A Comparison with Hartree-Fock and DFT Calculations, J. Rolke, Y. Zheng, C.E. Brion, S.J. Chakravorty, E.R. Davidson and I.E. McCarthy, Chem. Phys. 215, 191-205 (1997).
18. Nonlinear Substituent Effects on Multisubstituted Saturated Carbon and Radical Carbon: Anomeric, Geminal and Captodative Effects, E.R. Davidson, S. Chakravorty and J.J. Gajewski, New J. Chem. 21, 533-537 (1997).
19. Orbital momentum profiles and binding energy spectra for the complete valence shell of molecular fluorine. Zheng, Y., Brion, C.E. von Niessen, W., S.J. Chakravorty and E.R. Davidson Chemical Physics. (1996) 212, 269.
20. Simultaneous ionization and excitation of Helium atom and Hydrogen molecule studied by (e,2e) -spectroscopy, X Zheng, N.Lermer, B.R. Todd, N.M. Cann, C. E. Brion, E. Davidson, S Chakravorty, Int. Conf. Physics of Elect. And At. Collisions (1995).
21. Refinement of the Asymptotic Z Expansion for the Ground-State Correlation Energies of Atomic Ions, Chakravorty, S. J. and Davidson, E R. J. Phys Chem. 100, 6167 (1996).
22. A possible definition for Basis set superposition error, E Davidson, and S. Chakravorty. Chem. Phys. Lett. 217, 48(1994)
23. Atomic correlation energies for atomic ions with 3 to 18 electrons. Chakravorty, S.R. Gwaltney, E. R. Davidson and Farid Parpia and C.F. Fischer, Physical Review A 47, 3649 (1993)
24. The Water Dimer: correlation energy calculations S. Chakravorty and E.R. Davidson, Journal of Physical Chemistry 97, 6373 (1993)
25. The Transition metal - carbonyl bond, E.R. Davidson, K.L. Kunze, F.B.C. Machado, S.J. Chakravorty, Acc. Chem. Res. 1993, 26, 12, 628635
26. Test of the Hirschfeld Population Analyses, E.R. Davidson and S. Chakravorty, Theoret. Chim. Acta 83, 319 (1992)
27. Atomic Correlation Energies for two to ten electron atomic ions. E.R. Davidson, S.A. Hagstrom and S. Chakravorty C.F. Fischer and V. U. Meisar, Physical Review A 44 7071 (1991),
28. A Comparative Study of Density Functional Models to estimate Molecular Atomization Energies E. Clementi and Subhas J. Chakravorty, Journal of Chemical Physics & IBM Tech. Report 93 2591 (1990)
29. Soft Coulomb hole for the Hartree Fock model of Atoms. S. Chakravorty and E. Clementi, Physical Review A 39 2290 (1989) & IBM Tech Report KGN 156
30. Independent Electron Models: Hartree Fock for Many Electron Atoms Modern Techniques in Computational Chemistry Chapter 3 MOTECC 1989, MOTECC -1990, MOTECC -1991 S. Chakravorty, G. Corongiu, J.R. Flores, V. Sonnad, E. Clementi, V. Carravetta, and I. Cacelli.
31. Electron Correlation Energy from Hartree Fock type densities Modern Techniques in Computational Chemistry Chapter 15 MOTECC 1989 E. Clementi, S.J. Chakravorty, G. Corongiu and V. Carravetta.
32. ATOMSCF Program of MOTECC 89: Documentation Enrico Clementi and Subhas J. Chakravorty KGNMOL of MOTECC 89: Documentation G. Corongiu, E. Clementi, S. Chakravorty and P.L. Cristinziano.
33. Hylleraas Configuration Interaction method using Gaussian functions Chapter 5 MOTECC 1989, D. Frye, G.C. Lie, S.J. Chakravorty, A. Preiskorn, and E. Clementi.
34. Coulomb energy, total X ray scattering intensities and Average electron densities. S. Chakravorty and S.R. Gadre., Chemical Physics Letters 142, 205 (1987)
35. Nonlocal density approximation for exploring kinetic energy anisotropies S.R. Gadre, T. Koga and S. Chakravorty, Physical Review A 36, 4155 (1987).
36. Use of a nonlocal density approximation for the transformation of atomic electron density to electron momentum density S.R. Gadre and S. Chakravorty, Journal of Chemical Physics 86, 2224 (1987).
37. The average electron momentum density and rigorous bounds to average electron densities for atoms and molecules S.R. Gadre and S. Chakravorty, Chemical Physics Letters 132, 535 (1986).
38. Some novel characteristics of atomic information entropies. S.R. Gadre, S.B. Sears, S. Chakravorty and R.D. Bendale, Physical Review A 32, 2602 (1986)
39. Some rigorous inequalities among the Weizsacker correction and atomic <rn> and <pn>- values. S.R. Gadre and S. Chakravorty, Journal of Chemical Physics 84, 7051 (1986).
40. Compton profiles of atoms from electron densities via reciprocal form factors, S.R. Gadre and S. Chakravorty Proceedings of the Indian Academy of Sciences (Chemical Science) 96, 241 (1986).
41. Interconnections between atomic electron density and electron momentum density: leading and tail corrections S.R. Gadre and S. Chakravorty. Physical Review A 33, 1374 (1986).
42. The self-interaction correction to the local spin density model: Effect on the atomic momentum space properties S.R. Gadre and S. Chakravorty, Chemical Physics Letters 120, 101 (1985).
43. Hartree-Fock momentum expectation values for atoms and ions S.R. Gadre, S.P. Gejji and S. Chakravorty, Atomic Data and Nuclear Data Tables 28, 477 (1983).
44. On the monotonicity of atomic electron momentum density and shell structure of the radial momentum density S.R. Gadre, S. Chakravorty and R.K. Pathak, Journal of Chemical Physics 78, 4581 (1983)