Developing computational methods at the intersection of AI, biology, and human-relevant toxicology.
I work at the intersection of generative AI, machine learning, bioinformatics, and biomedical data science, developing computational approaches for complex biological and biomedical data.
My interests span toxicology and risk assessment, genomics, microbiome research, histopathology, and healthcare, with a focus on building models that enable biological translation, prediction, mechanistic insight, and data-driven scientific discovery.
Applying generative AI and machine learning to cross-domain biological translation, including in vitro → in vivo extrapolation.
I am also interested in developing AI-based virtual control group (VCG) approaches using treatment → control and control → control translation of clinical pathology profiles and carcinogenicity-study histopathology data, with the broader goal of advancing New Approach Methodologies (NAMs) and the 3Rs — Replacement, Reduction, and Refinement.
Developing computational approaches for high-dimensional biological and biomedical data spanning genomics, microbiome research, histopathology, toxicology, and healthcare.
Building reproducible bioinformatics, statistical, and machine-learning workflows for biological research, predictive modeling, and large-scale data analysis.
A generative AI framework for translating in vitro transcriptomic responses into synthetic in vivo profiles for toxicological applications.
The framework uses GAN-based cross-domain translation to model relationships between experimental systems and evaluate whether generated profiles preserve toxicologically relevant biological information.
Research themes
Generative AI · GANs · IVIVE · Transcriptomics · Toxicology · Risk Assessment
💻 Explore the repository → 📄 Read the publication in Toxicological Sciences →
A CVAE-GAN approach for deriving study-aligned synthetic controls and exploring AI-enabled virtual control groups in toxicology studies.
The work investigates generative modeling of control profiles with the broader objective of supporting approaches that can reduce reliance on concurrent control animals while preserving study-relevant biological information.
Research themes
Generative AI · CVAE-GAN · Virtual Control Groups · Toxicology · NAMs · 3Rs
💻 Explore the repository → 📄 Read the publication in Toxicological Sciences →
Python · R · SQL · Git
PyTorch · TensorFlow · scikit-learn
Generative Modeling · Predictive ML · Statistical Modeling
Genomics · Metagenomics · Metatranscriptomics
BLAST · QIIME2 · Network Analysis
Histopathology · Toxicology · Transcriptomics
Carcinogenicity Studies · High-Dimensional Data · Data Visualization
🧫 Network Modeling for Bacterial Communities
Microbial community network analysis using 16S sequencing data
SPIEC-EASI network construction and statistical analysis for studying relationships within bacterial communities.
Methods
Microbiome · 16S Sequencing · SPIEC-EASI · Network Analysis
🦠 SIR Model Simulation — COVID-19
Mathematical modeling of infectious-disease dynamics
SIR-based simulation with visualization of epidemic dynamics and herd-immunity behavior.
Methods
Epidemiology · Mathematical Modeling · Simulation · Visualization
🤖 Machine Learning Examples
Applied machine-learning implementations across several problem types
Examples include:
- Regression
- BERTopic
- Medical classification
- Naive Bayes
- General ML workflows
Interested in conversations and collaborations at the intersection of generative AI, bioinformatics, toxicology, genomics, and computational biomedical research.
🎓 Google Scholar: Mansi Chandra 🔗 LinkedIn: Mansi Chandra ✉️ Email: mchandra@ualr.edu
Computational approaches for translating biological data, modeling toxicity, and advancing human-relevant science.