Software & Projects
Research tools, pipelines, databases, and digital twin platforms
Portfolio
Research Projects & Software
A comprehensive portfolio of computational tools, ML pipelines, databases, and digital twin platforms developed by Sabyasachi Mohanty during his research career.
Molytics
Developed a machine-learning and AutoML platform for materials and property prediction, integrating formula featurization, dimensionality reduction, clustering, cluster-specific regression, unseen-data routing, confidence analysis, and scientific visualization.
Invention Disclosure: NUtech Ventures, UNL, 2025
CopolPred
Developed an LSTM and Siamese neural-network framework to predict monomer-pair reactivity ratios in copolymerization systems using curated experimental data and chemical descriptors.
CopolPred 2
Extended CopolPred with unsupervised clustering to group monomer pairs by chemical similarity and reactivity behavior before model training and prediction.
Polymer Copolymerization Prediction Using GAT-GNN
Designed graph-attention and graph-neural-network workflows to model polymer/copolymer sequence, monomer-structure relationships, and copolymerization-property prediction tasks.
Solid-State Electrolyte Property Prediction
Built scientific ML workflows for ionic conductivity and related materials-property prediction using clustering, regression, feature selection, and unseen-composition routing.
PolyForm: Polymer Digital Twin with Multiscale AI
Built a multiscale digital-twin platform connecting molecular descriptors, surrogate models, and process-scale prediction workflows for polymer systems.
1st Place, Husker AI Days Oral Presentation – 2026; People's Choice Award, UNL Research Days – 2026
SpadeTwin: Digital Twin for Chemical Reaction Engineering
Designed an AI-powered digital twin for chemical reaction engineering by coupling mechanistic reactor models with machine-learning surrogates for yield, selectivity, and process-optimization studies.
Invention Disclosure: NUtech Ventures, UNL, 2026
ProDrugDB: Prodrug Database
Compiled a curated prodrug database with chemical structures, activation pathways, target drugs, bioavailability profiles, clinical context, and searchable web functionality.
Community-Based Modeling / NSF MICRA
Modeled soil microbial ecosystems and Bacillus subtilis secondary-metabolite interactions to study microbial community structure, nutrient cycling, and drought-resilience mechanisms.
Funded: NSF MICRA 2024–Present and 2025
Pattern-Based Metabolic Changes
Integrated transcriptomic and metabolomic datasets to identify disease-specific metabolic signatures and prioritize drug-repurposing candidates for immune and inflammatory conditions.
Pharmacometric Modeling for Radiation Modulators
Collaborated on U.S. Army / DoD-associated pharmacodynamic modeling workflows for radiation-modulating drug candidates and medical-countermeasure research.
Funded: U.S. DoD / NSRI 2022–2024
Digital Twin for Immune Response
Designed multi-scale computational twins representing healthy and diseased tissue environments to simulate immune-cell behavior and drug-response outcomes.
AI-Driven Drug Discovery Using Generative Models
Developed generative-modeling workflows for de novo anticancer molecule design, compound screening, and candidate prioritization.
Brain Tumor Database (BTdb)
Built an integrated repository combining brain-tumor imaging, gene-expression profiles, and molecular-interaction networks from public data sources for biomarker and prognosis studies.
COMO Pipeline
Pipeline for multi-omics data integration in metabolic modeling and drug discovery, published in Briefings in Bioinformatics.
Optrecon
Optimization-based framework for automated genome-scale model reconstruction upon resolving thermodynamically infeasible cycles.
Invention Disclosure: NUtech Ventures, UNL, 2025
IP
Invention Disclosures
Invention disclosures submitted to NUtech Ventures, the technology commercialization office of the University of Nebraska-Lincoln.
SPADETWIN
NUtech Ventures, University of Nebraska-Lincoln
Molytics
NUtech Ventures, University of Nebraska-Lincoln
Optrecon
NUtech Ventures, University of Nebraska-Lincoln
Development
Software Development Resources
GitHub Repository
All publicly available code, pipelines, and tools maintained by Sabyasachi Mohanty.
codeGitHub: sm1216Technology Stack
Primary tools and languages used in software development:
Live Applications
Web applications and platforms developed for research:
AI-driven clustering and AutoML framework for scientific property prediction and materials discovery.
Siamese-LSTM framework for copolymer reactivity-ratio prediction and model-guided copolymer design.
Curated prodrug database with chemical structures, activation pathways, and clinical context.