About

Shivaprasad Patil, PhD

Associate Director, Bioinformatics & Predictive AI

Scientist working on AI models for drug discovery using genomics and multi-omics data. I focus on translating biological signals into actionable insights for R&D teams.

GitHub Google Scholar LinkedIn Email Request CV
Shivaprasad Patil portrait

Current Focus

  • Generative and predictive AI for drug discovery
  • Multi-omics integration across bulk, single-cell, and spatial data
  • Translational biomarkers for patient stratification

About

I am an Associate Director of Bioinformatics and Predictive AI at AstraZeneca with over a decade of experience in computational biology and machine learning. I work at the intersection of genomics, data science, and drug development to turn high-dimensional biological data into actionable insights.

  • Background: PhD in Biostatistics and Biomedicine, Technische Universitat Dresden.
  • Research interests: AI/ML, genomics, multi-omics, translational biomarker discovery.
  • Current focus: robust AI models for target validation and safety.

Research and Projects

Selected projects that showcase my work on scalable AI systems and multi-omics integration.

MultiOmicsBind infographic

MultiOmicsBind

Deep learning framework inspired by ImageBind to align genomics, transcriptomics, proteomics, and metabolomics for integrated insights.

Methods: multimodal learning, representation alignment

View on GitHub
ScAdver infographic

ScAdver

Adversarial batch correction for single-cell RNA-seq that preserves biology while removing technical variation across protocols.

Methods: adversarial learning, domain adaptation, scRNA-seq

View on GitHub
BulkRNA-Agent infographic

BulkRNA-Agent

AI-powered transcriptomics analysis with dual LLM reasoning and biomedical models, covering QC through enrichment analysis.

Methods: LLMs, RNA-seq, differential expression, enrichment

View on GitHub
OmicsFormer infographic

OmicsFormer

Transformer-based framework for multi-omics integration with robust handling of missing modalities, batch effects, and cross-study generalization.

Methods: transformers, batch correction, multi-omics

View on GitHub

View all projects

Publications

Peer-reviewed articles and conference proceedings.

Browse publications

Skills

AI and Machine Learning

Deep learning Generative models Representation learning Model interpretability Causal inference

Genomics and Bioinformatics

Bulk RNA-seq Single-cell omics Spatial transcriptomics Proteomics Multi-omics integration

Programming and Tools

Python R PyTorch TensorFlow Nextflow Docker

Experience

Associate Director, Predictive AI & Bioinformatics

10/2024 - Present · AstraZeneca · Barcelona

  • Developing predictive AI and bioinformatics pipelines to translate omics and imaging data into mechanistic insights.
  • Building multi-omics workflows to refine toxicity predictions and accelerate candidate prioritization.
  • Designing AI models to extract predictive morphology from Cell Painting for safety decisions.
  • Advancing proteomics-driven models to improve compound selection in early discovery.

Scientific Investigator - Human Genetics & Genomics

07/2022 - 09/2024 · GSK · Heidelberg

  • Integrated bulk and single-cell multi-omics to uncover biological mechanisms across programs.
  • Performed post-hoc analyses of clinical trial data to inform precision medicine strategies.
  • Built translational modeling frameworks linking iPSC-derived Alzheimer’s models with multi-omics data to characterize disease progression.
  • Identified molecular subtypes and translational signatures that enabled indication expansion.

Bioinformatician & ML Researcher

06/2019 - 06/2022 · National Center for Radiation Research in Oncology - University Hospital Dresden

  • Predicted patient prognosis in head and neck squamous cell carcinoma across 8 cohorts (1500 patients).
  • Applied statistical and machine learning methods to identify biomarkers for personalized radiation oncology.
  • Enabled patient selection for prospective clinical trials with multidisciplinary teams.
  • Delivered 50+ poster and oral presentations and served as a reviewer for Radiotherapy & Oncology.

Bioinformatics Research Associate

08/2017 - 09/2018 · Katholieke Universiteit Leuven · Belgium

  • Built a pipeline for viral metagenomic analysis and assembled a catalog of 3000 viral genes and proteins.
  • Identified biomarkers for early diagnosis of liver fibrosis in a cohort of 200 patients.

Bioinformatician

2015 - 07/2017 · Institute of Genomics & Integrative Biology · New Delhi, India

  • Identified and annotated 60,000+ novel circular RNAs across 38 datasets.
  • Part of a multi-center collaborative project with 280 clinicians and researchers across 60 centers, advancing exome sequencing for rare disease.
  • Advanced variant prioritization and non-coding RNA studies for gene regulation insights.

Education

  • PhD, Biostatistics and Biomedicine - Technische Universitat Dresden (2019-2022)
  • Integrated BS-MS, Life Sciences - IISER Mohali (2010-2015)